I agree that some degree of "system mastery" is required for feats, but that certainly isn't unique to the playtest, nor is it unique to D&D. That said, the final 5E feat system has its own host of problems, which are more egregious than an evolution of the 3ed/Next framework.
With all games, a balance must be struck between intuitive rules and requiring the player to do "homework". As games go, D&D is rather complex, but its advantage is that you don't need to know everything up front. That said, at some point, a player is going to have to open the book and read it. Certainly the debate is where the balance lies.
I also didn't like that the playtest turned a lot of melee class abilities into feats, or at least a weird hybrid of (maneuvers). I'm fine with feats being "required", but I absolutely agree that there needs to be choices within the system that are easily understood so a player won't be "punished" for not having the same level of mastery as another player has (which absolutely does happen now in 5E).
I guess if I'm being honest, my negative reaction to the playtest packet 7 feats is just distaste for having to learn a new and different system for the trivial purpose of satisfying my curiosity. There's nothing inherently bad about the feats being smaller and more numerous.
What I'd insist on, though, is that featifying the design makes it a lot more kludgy. The feats at 1, 3, 6, and 9 being baseline, as you say, shifts the system mastery balance towards players having to do more homework. The designers seemed to recognize that and added specialties--collections of 4 feats with a couple paragraphs of fluff--which are totally optional but function almost like universal subclasses. I bet there are tons of traps in them for people without sufficient system mastery (though I admit I haven't looked into it). It seems like a poor compromise.
As an extreme example of featified kludge, the monk's martial arts feature gives them
the martial arts feat. What does this central class feature do? look it up in the feats doc. Moreover, any PC can take the martial arts feat at 1st level, even a wizard (though its a pretty bad option for wizards, since they don't get an attack bonus). The martial arts die never improves and the only unique advantages monks get with unarmed attacks is that they are magical and have some on hit ki abilities (depending on the subclass). Granted, you already stated that you aren't a fan of melee class abilities being turned into feats, but I felt like that one bore mentioning.
I generally like feats and the character optimization minigame. But reading through the old playtest packet has given me a sense that I also like 5e's baseline expectation that players will opt out of large parts of it. And, apart from GWM and it's ilk being in the game (if they are allowed), vanilla PHB PCs aren't punished for that. Maybe if the playtest's 1, 3, 6, and 9 feats were written as standard PC features with an optional rule allowing players to swap them out, they would seem more appealing to me. They are kind of set up that way anyway with a suggested specialty in every class description, so it could just be a rules presentation issue.