Where exacly I said there is a "big gap"? Reread all my posts if you want, the only thing I write was "below average" (about 500 times), you misinterpreted because you can't say "yes, you're right, a bard group will stay a tiny line below average in ofense/defense at start lvls. Not significant in fact, we talking about 1 or 2 points, and this will gone later, but will be beneath at lvls 1 and 2." This reminds me Leon Festinger about cognitive dissonance.
That reminds me about pretentious comments that don't actually contribute to the debate. ;-)
You are still arguing from an average that has not been established. It's based on your preconceived notion of what the average actually is and is entirely subjective as opposed to objective. All you've really done is determine an arbitrary baseline as "the average". There is no gap at all when it comes to hit points because d8 is the most common hit die. That's a fact. There is no gap at all when it comes to damage because TWF is on the higher end of damage options at those levels (damage gap occurs at higher levels and impacts bards more later).
What you are calling below average is your comparison of a character with 16 AC instead of 15 or 14 and ignoring the other characters in the standard group who also have 15 AC or those with 12 or 13 AC. An average includes the lower with the higher. That's why it's an average. A person cannot point to the slightly higher AC options and consider that the average just because it's higher than what most bards are at for AC.
Your idea of average is way off base.
The fact is that to prove your point you're over optimizing bards, but not doing the same with the other classes. Assuming the standard for bards are Dex 16, Con 14 for the HP, but not Cha 16 that is the primary stat for bards? Weird.
What's weird is the assumption you have for melee oriented bards to need one more inspiration die because that's all they'll get out of 16 CHA when they are focusing more on melee than spell DC's. Taking things like healing word and heroism don't require a spell DC. That's no different than any other spell casting class who is focused on combat. DEX is important to bards because it prevents damage based on the save type, contributes to a lot of appropriate ability checks, and covers both ranged and melee attacks along with AC.
Focusing on some defensive qualities is just common sense when the group knows that all the defense and all the offense will be handled by bards in an all bard party. Your argument requires that a group willfully neglects those options and hopes for the best.
It's also not over optimizing bards to look at a standard ability score for combat for the class and possible feat for one or two of them. This is particularly true when your examples below include things like 16 DEX and 16 WIS for a monk to make that 16 AC that you think they would have.
You're saying that a bard group will have TWF, but some of them will get moderately armor for better defense, so, they will have better AC OR better damage, not both.
The armor is only relevant for shield use. This is the same as every other class with a melee aspect at those levels. They either take a shield or they don't. The AC of 15 for a bard is still the same as the AC for a TWF ranger, for example.
You are applying a principle to one class as an issue when it's the same for all. The bard can maintain TWF and spell casting as per my previous post, or if they do go with a shield proficiency they can do the same with sword'n'board upgrading to a d8 weapon and a swap to match most other users of that style. There isn't a difference between a bard and most classes when it comes to AC and damage.
about 95%, well, the OP party plan doesn't include any one that is thinking to get moderately armor, because later with 3 valor bards and 25 bard inspiration, there is much better things to do with that feat or extra stats for being a half-elf. Search for all bards buildings in foruns to check this statistic.
"95%" is obviously an arbitrary number you pulled out of your hat with nothing to back it up. It's a made up statistic.
Like I said earlier, it's for the shield bonus and not the AC. The AC is still similar to most AC and the upgrade is for lore bards who want better AC by using the shield or taking 14 DEX with no intent to increase their DEX. The valor bards wouldn't take it, but it's a good option for bards planning on going lore for defense and usually mentioned as such in guides.
What some internet builds show in their builds does not represent builds in an all bard party.
looking about builds in character optimization area, the fact about defense:
barb, druids, fighters, monks, palis, rangers, and half clerics will have better AC and/or HP than bards. Maybe just 1 point of difference, but is a difference.
rogue, warlocks and half clerics will have the same
wizards and sorceres will be lower.
The barbarian likely has worse armor or the same armor. They build STR, not DEX, and tend to focus on big weapon styles. What they have is 4 more hp at 1st level and 6 more hp at 2nd level. That's not even close to an issue given the amount of healing available in an all bard party.
Druids are likely to have worse armor or the same armor. They do not normally build DEX either because it doesn't help them when they shape shift and they can focus on WIS and shillelagh for an attack ability score. They have the same hit points although they can create a pool of hit points with a worse AC and damage options to go with it. Druids are almost exactly the same in AC and hp with worse damage options for the typical druid.
Fighters are no argument there. The average party isn't all fighters and paladins, however. ;-)
Monks typically have 15 AC from what I've seen. It's possible to push them to 16 but that seems to be outside of the norm, although the counter to that is it's possible for the lore bard to use a shield for better AC and the same hp. AC and hp are still almost exactly the same but the monks do have the damage with martial arts and a quarterstaff.
Paladins are no arguments there.
Rangers have the same AC as bards with the only exception being that they have the shield proficiency for free. Most rangers are archers or TWF. They might go for more of an AC build with the defense or dueling fighting styles at second level, but archery seems the most typical to me so far. They have 2 more hp at 1st level and 3 more hp at 2nd level. Just like the barbarians, that is not an issue given the large number of spell slots available to the group.
Clerics who take domains without the armor upgrade generally have the same armor but not necessarily DEX. They do typically use shields in my experience and would have better AC with the same hit points. An all cleric party can work well.
Rogues have the same hit points and AC. No argument there.
Warlocks start out with the same hit points but only those going blade pact might look at more defensive options. In an all warlock party I would expect at few to follow the same defensive options as bards in an all bard party. Normally I don't see warlocks going for that higher DEX investment but they might add some medium armor and shield as well. What I see is less combat oriented warlocks but they do have the same options at first. Eldritch blast isn't as good as TWF either way when it comes to damage.
About offense:
barbs, fighters, monks, pali, rangers, rogues, 1/2 druids, 1/3 clerics will have better.
1/2 druids, 2/3 clerics, bladelocks and bladesingers will be equal
sorcerers, others warlocks and wizards will be lower.
At those levels, barbarians barely have an offensive benefit over bards. All they have is rage when it can be maintained for that +2 bonus to damage. The additional disclaimer is bardic inspiration and spell power.
Fighters are no argument there. The fighting style benefits are available immediately. The additional disclaimer is bardic inspiration and spell power.
Monks are no argument there. Martial arts allows for using a staff two handed with DEX mod and then still getting an unarmed attack in as a bonus action. Solid damage to go with the same AC and hp. The additional disclaimer is bardic inspiration and spell power.
Paladins don't have benefits over the bard at 1st level but pick them up at 2nd level with the fighting style. They also have nice buffs (which bards don't, typically) in the spell list.
Rangers don't have benefits over the bard at 1st level, but pick them up at 2nd level with the fighting style and possibly a bit with hunter's mark.
Rogues have sneak attack damage, which is one of the better options, like monk. The additional disclaimer is bardic inspiration and spell power.
Druids don't have benefits over the bard at 1st level. Moon druids have some options at 2nd level but those are also part time options and druids can turn into spectators instead of participants. They end up dropping the forms for other actions (like casting spells) and they lose hit points (and forms) because of lower AC's. It's an option that looks better on paper than in practice, works great at low levels and epic levels, and not so much the rest of the time.
Clerics don't have better offensive options than bards at those levels. Simply gaining martial weapon proficiency isn't a bonus over TWF at low levels. There are a few minor abilities on limited uses per day but something like 2 (maybe 3) bonus action attacks from a war domain cleric are almost pointless except on some focused builds to make use of it. Given that most clerics do typically use shields in my experience they normally do less damage than bards. They do have some nice buff spells, at least. Clerics are not better or equal, in general.
Warlocks planning on going blade pact are not as good as bards when it comes to offense at low levels. The issue is in the simple weapons proficiency and planning for the blade weapon. Bards can afford to go DEX and dual wield short swords. Warlocks don't have short sword proficiency. If they want to use DEX they are dropping to daggers, and if they want to use different weapons they need to invest in STR and do something else for defense. Blade pact warlocks take planning to do well later and give up on those first couple of levels. Bardic inspiration also favors the bards.
If you count, will see the "below average thing".
please, don't make me do overly optimized characters to low lvls of each class to prove that.
Counting indicates you are applying a median average instead of a mean average or mode average. It doesn't work because there are no actual numbers applied from which to base where the bard fits above or below such a median. What you are attempting to do is stack rank them based on generalization (and some misconceptions) stemming from subjective opinion. Your argument still stems from your preconception of what the average is.
Optimized builds are not indicative of any average, regardless; they focus on extremes.