My perspective is that forum conversations often are so narrowly focused on 1%er problems that they tend toward or can devolve into the ludicrous.
Not to say that it's irrelevant how BiS builds compare in various ways, but there is such a thing as less than BiS builds, way less than BiS builds, "this common armor is better than this other common armor" builds, etc.
When mucking about with classes, can't just ignore that most players aren't decked out with any tokens they want. I would even go so far as to say that the first priority with balance should always be no tokens besides some assumed common weapons and armor when it comes to such things as classes, though "first priority" may not be that meaningful a concept. While different classes are going to do better or worse at different rarity levels, non-modified character card and .1%er builds are both places I could see effort put into as all of the in between is extremely hard to analyze.
Due to the inability to math everything, can run into the problem of deciding whether a class is mathematically better. How much do saves matter, for instance? Wizards may be math-sucky in many ways, but is the ability to ignore STR factored in to how optimizing builds leads to value?
Btw, characters do compete with each other. A recent blog post of mine was specifically about how TD and RPG PC builds are typically engaged in a player-side competition. Not always with other players, though often with other players. This has always been the case in my home RPG play (perhaps yet another reason I tend to enjoy convention one-shots more than home campaign play). But, it is more visible in the sort of gaming I engaged in at GC - TD and HoR, thus why I've been thinking more about it recently.
Players have expectations in their heads as to what is good, better, worse. They may be mathematically based or a special form of insanity, but they exist. If I run URs in a build, I expect it to be more better than a rare build. I expect it to be Hardcore+ viable, and so forth, i.e. there is context when it comes to builds (even ignoring other PCs' existence).
Part of that context is comparing against other people's builds. If I consistently do. let's say, 30 damage a round with my most awesomest build and the other players consistently do 50, then that affects my perception of my PC's value because, while I'm frequently happy to be sidekicking, lot's of RPGers and, assumedly, lots of TDers like getting equal billing.
If I consistently do 20 less damage a round and also heal or also give everyone else +4/+4 or get three free rares, I mean, help more with puzzles than others, then, sure, that seems fair. If I consistently do 20 less damage and only have as a differentiator a better AC than someone, well, not feeling fair.
Now, another aspect of PC builds in both TD and RPGs that I've felt strongly due to recent play of both is that the more you can improve a build the less important teamwork is. With RPG play, especially the zero to hero style of play, consider how much more dependent PCs are at low levels than when they can start teleporting.
With TD, consider that out of the starter pack play does require a lot of party intrareliance where BiS play doesn't. I'm not so far removed from casual play (where others seem to have no concept of it) that I don't remember what it was like where you didn't decide to resurrect the monster because you were bored after one-rounding it or where you couldn't just have the AC 47 player solo a fight while everyone else chitchatted. Weakness creates dependency, corollary being strength reduces dependency.
Also, there's far too little attention paid to how TD isn't just all about combat. Puzzle damage and push damage from puzzle rooms does actually exist, even if it's meaningless for 1%ers who are on their third run of a dungeon.
Then, it also seems clear that the definiton of success is very different for different players. I don't feel a need to win every combat. It would be boring if I knew I could win every fight. We came up 3hp short of winning a fight at GC and that made for a vastly better story than if we Prismatic Spray locked the monster until it ceased breathing. Mike may believe all that matters is damage output, but, across a variety of levels of play and different party configurations (such as shorthanded parties), healing/damage prevention/death prevention is far more relevant because success isn't always clean card but can also be "I got one slide in in room 7 before my soul was rent".
Really rangering off topic, but I'd rate TD a better game when played with a lack of collection. On the other hand, I'm a collector with games and I enjoy having expensive tokens even if it means the play experience is worsened.
Trying to get back on topic, are monk and ranger overpowered when they have +40 damage? What if they have +5? Is druid a tokenless god? Is fighter always the worst class in TD because it has no meaningful feature? What of the difference between playing 4th level and 5th level PCs because, I know this sounds bizarre, but some people actually don't have level-bumpers?
These questions do actually matter because players do actually care about how powerful they are not just in comparison to the challenges but in relation to other people's builds and in relation to theoretical builds in their deranged, er, I mean, gamer minds.