Leatherhead
Possibly a Idiot.
Because Someone had to do it. 
Actually it's because the lore wizard is too good of a concept to pass over due to problems, but whatever.
What do I mean by that?
Firstly, it's a generalist Wizard, something noticeably lacking from 5e.
Secondly, it's one of the few cases where a character can be smarter than the person playing it, which is really needed for a class like the Wizard, who is supposed to boast lots of brainmeats.
Thirdly, it interacts with the Wizards spellbook in a meaningful way, even during combat, which is really something no other wizard even attempts to do.
Now, lets talk about the subclass and its abilities.
Lore Master
This ability does two things:
Grants to essentially expertise in whatever INT based lore skills you have.
Lets you use INT on your initiative check.
The first is thematic, and at first glace it's a ribbon, but in actuality it's the backbone of the subclass. The general idea being: You see a monster, you get a lore check on the monster, then you exploit the monster's weaknesses. This is what allows the character to be smarter than the player, because lots of people don't know everything there is to know about D&D, but this lore master is a nerd who reads books so they know some things (well sometimes, they can still fail the check). Of course, it also works out of combat too, which is nice. Some would argue this steps on the toes of the Rogue, but really the ship of giving out extra bonuses to some skills has long since sailed away from Rogueport.
The second bit, is largely unnecessary. It's vaguely reminiscent of mechanics from previous editions that allowed players to use INT in place of DEX, but even those mechanics tended to stop short of applying to initiative checks. So it's mostly unprecedented (at least for anything looking like a wizard)
The verdict, scrapping the second part is fine. But the first bit is what drives everything.
Spell Secrets
This ability also does two things, both of which are related to exploiting the weaknesses of creatures.
Firstly, it lets you change the damage type of a level 1 or greater spell to another damage type.
Secondly, it lets you change the saving throw of a level 1 or greater spell to another saving throw.
The first bit causes concern for a few reasons. But really, they are mostly knee-jerk reactions.
Only 20 monsters in the MM have a vulnerability, meaning it isn't going to be a reliable source of extra damage, only a reliable source of doing normal DPR despite resistances. A common discussion point is “Force damage is too OP!” Which dismisses the fact that you don't even need force or even radiant damage to bypass every resistance in the game. The other common complaint is that Sorcerers exist, which isn't really legitimate reason to stop a Wizard subclass from doing things, only a reason to take a long hard look at the Sorcerer, much like how they revised the Ranger.
Other people say that being able to swap damage types on every spell you memorize when you cast them is too much, and well, that's a fair point. The way spellcasting works in 5e already lets you hedge your bets by memorizing a handful of damage spells and just casting them in higher level spell slots to help make up for a bad pick that day. Making it so you only need to memorize one damage spell then swap it out as you cast it, allows the player to focus all of those freed up spell slots on whatever else they can possibly want, which has huge implications. Limiting this part of the ability to a short rest isn't unwarranted, especially to prevent dips from the elemental themed subclasses who would love for everything to be one elemental damage type, but limiting it any farther is too much.
The second bit however, can be outright broken. Using STR or DEX saves on Hold Person, for example, can permanently lock down a target with no chance of escaping. The very concept of the ability should be higher level, if not a daily resource. And needs some way to avoid the hold person problem (like reverting to the normal save on subsequent rounds). That doesn't mean it should go away, it does fit the idea of “hitting the target where it hurts” but it does need a nerf of some kind.
Alchemical Casting.
Aside from having an esoteric name, this is basically metamagic via convergent mechanics. Burning spell slots instead of points to buff up your spells.
It has three parts:
1st level slots grant extra “AoE” force damage. I say “AoE” because it has some interesting interactions with Magic Missile of all things, allowing you to have a 100% accurate multi-minion-swater with IFF technology. Granted, 11d4+11+22d10 total damage sounds like a mountains worth, but you are only dealing 15~ average damage to any individual target, so good for minion cleanup, not really good for anything else.
2nd level slots give you a range of 1 mile, for reasons. I got nothing. This is just absurd. You could use this to take out an entire army at once.
3rd level slots give you a +2 bonus to your saving throw DC. At first blush this is unnecessary given Spell Secrets. But it can be used in conjunction with that ability to make sure the target isn't making that saving throw.
To be honest, this is completely unnecessary. It theoretically represents the wizard having mastery of magic, which is kind of the theme of the entire subclass, but all the other abilities cover that nicely, and all this does is cause problems it seems. Some problems being worse than others, obviously. For a quick replacement, you could pop the second bit of Spell secrets up to this level, after you fix the problems of that ability.
Prodigious Memory
This is hands down the most wizardly ability ever. And one of the main reasons I would like to see this subclass see the light of day. And I haven't seen much in the way of complaints about this, though that might be due to people freaking out over the other aspects.
You swap one of your prepared spells out as a bonus action, once per rest. It's so simple, yet so perfectly intuitive that you have to wonder if all wizards shouldn't have this.
Master of Magic.
You can cast spell of whatever class once per day, providing you have the spell slot for it.
Some people have said this is more powerful than Wish, I disagree, it's more like Limited Wish of previous editions, but with access to higher level spells. It nets you access to three spells most other wizards won't have even limited access too: Mass Heal, Storm of Vengeance, and True Resurrection.
Realistically though, this ability is going to be used to grab a spell from a spell list your Wizard doesn't have, on the rare occasion that it's absolutely necessary, such as casting Resurrection on the Cleric. And it doesn't need to have access to spells higher than 7th level for that kind of thing to still be true. Granted, most people don't even play at a level where such a thing will even be a consideration, but it's nice to future proof your work.
Normally I wouldn't do an entire thread about one subclass like this, but this one is generating lots of discussion, so feel free to discuss.

Actually it's because the lore wizard is too good of a concept to pass over due to problems, but whatever.
What do I mean by that?
Firstly, it's a generalist Wizard, something noticeably lacking from 5e.
Secondly, it's one of the few cases where a character can be smarter than the person playing it, which is really needed for a class like the Wizard, who is supposed to boast lots of brainmeats.
Thirdly, it interacts with the Wizards spellbook in a meaningful way, even during combat, which is really something no other wizard even attempts to do.
Now, lets talk about the subclass and its abilities.
Lore Master
This ability does two things:
Grants to essentially expertise in whatever INT based lore skills you have.
Lets you use INT on your initiative check.
The first is thematic, and at first glace it's a ribbon, but in actuality it's the backbone of the subclass. The general idea being: You see a monster, you get a lore check on the monster, then you exploit the monster's weaknesses. This is what allows the character to be smarter than the player, because lots of people don't know everything there is to know about D&D, but this lore master is a nerd who reads books so they know some things (well sometimes, they can still fail the check). Of course, it also works out of combat too, which is nice. Some would argue this steps on the toes of the Rogue, but really the ship of giving out extra bonuses to some skills has long since sailed away from Rogueport.
The second bit, is largely unnecessary. It's vaguely reminiscent of mechanics from previous editions that allowed players to use INT in place of DEX, but even those mechanics tended to stop short of applying to initiative checks. So it's mostly unprecedented (at least for anything looking like a wizard)
The verdict, scrapping the second part is fine. But the first bit is what drives everything.
Spell Secrets
This ability also does two things, both of which are related to exploiting the weaknesses of creatures.
Firstly, it lets you change the damage type of a level 1 or greater spell to another damage type.
Secondly, it lets you change the saving throw of a level 1 or greater spell to another saving throw.
The first bit causes concern for a few reasons. But really, they are mostly knee-jerk reactions.
Only 20 monsters in the MM have a vulnerability, meaning it isn't going to be a reliable source of extra damage, only a reliable source of doing normal DPR despite resistances. A common discussion point is “Force damage is too OP!” Which dismisses the fact that you don't even need force or even radiant damage to bypass every resistance in the game. The other common complaint is that Sorcerers exist, which isn't really legitimate reason to stop a Wizard subclass from doing things, only a reason to take a long hard look at the Sorcerer, much like how they revised the Ranger.
Other people say that being able to swap damage types on every spell you memorize when you cast them is too much, and well, that's a fair point. The way spellcasting works in 5e already lets you hedge your bets by memorizing a handful of damage spells and just casting them in higher level spell slots to help make up for a bad pick that day. Making it so you only need to memorize one damage spell then swap it out as you cast it, allows the player to focus all of those freed up spell slots on whatever else they can possibly want, which has huge implications. Limiting this part of the ability to a short rest isn't unwarranted, especially to prevent dips from the elemental themed subclasses who would love for everything to be one elemental damage type, but limiting it any farther is too much.
The second bit however, can be outright broken. Using STR or DEX saves on Hold Person, for example, can permanently lock down a target with no chance of escaping. The very concept of the ability should be higher level, if not a daily resource. And needs some way to avoid the hold person problem (like reverting to the normal save on subsequent rounds). That doesn't mean it should go away, it does fit the idea of “hitting the target where it hurts” but it does need a nerf of some kind.
Alchemical Casting.
Aside from having an esoteric name, this is basically metamagic via convergent mechanics. Burning spell slots instead of points to buff up your spells.
It has three parts:
1st level slots grant extra “AoE” force damage. I say “AoE” because it has some interesting interactions with Magic Missile of all things, allowing you to have a 100% accurate multi-minion-swater with IFF technology. Granted, 11d4+11+22d10 total damage sounds like a mountains worth, but you are only dealing 15~ average damage to any individual target, so good for minion cleanup, not really good for anything else.
2nd level slots give you a range of 1 mile, for reasons. I got nothing. This is just absurd. You could use this to take out an entire army at once.
3rd level slots give you a +2 bonus to your saving throw DC. At first blush this is unnecessary given Spell Secrets. But it can be used in conjunction with that ability to make sure the target isn't making that saving throw.
To be honest, this is completely unnecessary. It theoretically represents the wizard having mastery of magic, which is kind of the theme of the entire subclass, but all the other abilities cover that nicely, and all this does is cause problems it seems. Some problems being worse than others, obviously. For a quick replacement, you could pop the second bit of Spell secrets up to this level, after you fix the problems of that ability.
Prodigious Memory
This is hands down the most wizardly ability ever. And one of the main reasons I would like to see this subclass see the light of day. And I haven't seen much in the way of complaints about this, though that might be due to people freaking out over the other aspects.
You swap one of your prepared spells out as a bonus action, once per rest. It's so simple, yet so perfectly intuitive that you have to wonder if all wizards shouldn't have this.
Master of Magic.
You can cast spell of whatever class once per day, providing you have the spell slot for it.
Some people have said this is more powerful than Wish, I disagree, it's more like Limited Wish of previous editions, but with access to higher level spells. It nets you access to three spells most other wizards won't have even limited access too: Mass Heal, Storm of Vengeance, and True Resurrection.
Realistically though, this ability is going to be used to grab a spell from a spell list your Wizard doesn't have, on the rare occasion that it's absolutely necessary, such as casting Resurrection on the Cleric. And it doesn't need to have access to spells higher than 7th level for that kind of thing to still be true. Granted, most people don't even play at a level where such a thing will even be a consideration, but it's nice to future proof your work.
Normally I wouldn't do an entire thread about one subclass like this, but this one is generating lots of discussion, so feel free to discuss.