Spell Versatility and the Bard raises my eyebrows. So now for the cost of a single Known Spell the Bard can bi-daily cast Awaken or Raise Dead...
or any of these spells: (SRD Copy)
Animate Objects
Awaken
Dominate Person
Dream
Geas
Greater Restoration
Hold Monster
Legend Lore
Mass Cure Wounds
Mislead
Modify Memory
Planar Binding
Raise Dead
Scrying
Seeming
Teleportation Circle
The Bard has the arguably the best Utility Spell list, which was balanced by having a max of 22 spells known, so selecting spells was a bit of a Sophie’s Choice conundrum...with UA rules as proposed, does the Bard need more than 18 Spells Known?
A player with a Wizard by RAW could pay up to a max of 4000 gp and 160 hours (20 Long Rests) to have 16 Fifth level spells in their diary.
Swapping Command for Charm Person, not big deal....but Awaken and Raise Dead, (& other 5th + LVL spells) are campaign changing events.
Balance wise, a ‘create a friendly NPC’ spell should probably always cost a Known Spell slot.
Do others agree, or is that giant really a windmill?
It's a long post. I apologize in advance.
The bard is actually the class I'm looking at the most closely. If any class can exploit this mechanic it's going to be the bard.
The conundrum is still that bards are never going to have any more 5th-level spells at any given time than they learn from leveling up. In my case that tend to be limited to raise dead and greater restoration already.
We're still back to the wizard having 2 or 4 spells in his book to prep regardless of the cost of adding more. At any given point in time that wizard will have 14 or 15 spells prepped to my bard's 12 or 14 spells known. This is at 9th and 10th levels respectively, and 10th level is actually a sweet spot for bards.
Spells known compared to prepped isn't as much of a challenge for bards but they can get pulled into a few different directions so spreads a bit thin covering healing and control. I focus on healing as you can see in my typical spell selection. Dropping raise dead is unlikely unless I add revivify, which would have cost the same number of spell secret anyway (it's not a bard spell), because waiting a day to raise a party member is a pita. Party members don't normally die during downtime so it's not relevant.
I can already cast raise dead any day of the week on my bards. It can't be more campaign breaking than that.
The problem is you need to actually apply those spells to downtime and determine what the benefits are. Awaken takes 8 hours to cast and costs a 1000gp agate. That's limited by the DM who controls the supply of 1000gp agates, for starters, and doing it 5 times cost more than your hypothetical scribing costs. If it's campaign breaking it's campaign breaking regardless of swapping because a bard can already take it and a druid can already swap it in or out at will.
The ability to swap in or out any spells that might be problematic would need to be spells that don't also exist on the cleric and druid lists. It's not possible to point to a spell as being broken or game changing by being able to swap it in if we already have classes that can already do that demonstrating it's not game breaking.
Compared to the wizard, bards do line up spells known with spells prepped rather well. What happens at 9th level is the wizard learns 2 spells and the bard learns 1 spell, and can trade 1 spell up. That demonstrates the wizard is gaining while the bard is trading. At 10th level the bard doesn't gain any regular spells known and gains secrets instead (the reason for the sweet spot and not truly representative overall). Both can be used for 5th level spells but I usually use 1 for a 5th-level spell and 1 for a 4th-level or lower level spell. I find it's better to add spells that I can apply to more available slots than spells where I have minimal slots. That's one of the drawbacks to known spells. The wizard learns 2 more 5th level spells in his book.
On the surface that's similar. It's 2-4 5th-level spells for the bard and 4 5th-level spells for the wizard. Given that both of these classes are geared towards versatility it's also pretty much expected. The wizard also has tradition features to enhance spells at 10th level that the bard doesn't get. Every tradition feature before that enhances spell casting that the bard doesn't get. The wizard has 5 levels worth of spell slots the bard also does not get via arcane recovery. It can be spent on an extra 5th-level slot or multiple lower level slots but it's more spells per day than the bard can cast. The wizard has a much better ritual casting mechanic than the bard gets.
The best the bard can pull off in 5th-level spells known is the minimum the wizard can pull off in spells added to the spell-book, and the wizard can cast more of them, and better, at a peak level for bards. That peak level is just before the wizard break away where they leave every other arcane spell caster behind because it's the slow down period for spells-known casters. That's about as close as spells-known casters can get. The bard advantage is they get skill bonuses and bardic inspiration instead. That's debatable on whether the extra spell power is less than, equal to, or greater than skills and inspiration but those have no impact on the spell versatility feature.
The only benefit for spell versatility I see in that list is taking a day to add awaken, scrying, or teleportation circle if there's time, and then taking another day to trade back for the combat day. Simply swapping one puts the bard down an important spell.
The important thing to consider here is that players aren't actually playing or adventuring during downtime. The number of spells known doesn't actually change and returning to play even if a person changed the entire spell list only means the wizard and bard would be compared as if the bard had selected that entire other spell list in the first place. It's not actually a significant mechanical advantage during gameplay from what I'm seeing so far. Downtime requires the assumption the wizard doesn't take some downtime spells as well, which wizards do, and they have a solid selection.
When it comes to spells the wizard has the advantage at every turn.
If you want to push the cost of scribing scrolls, then the minimum the wizard can do is going up even though it was already better than the bard's maximum. That's the gravy effect.
It was a long post, but thank you to anyone who read it all. 
