I have a great deal of interest in making D&D easier to learn and teach, because that helps grow the hobby, and the rules we have (even in 5e!) are all too often byzantine and excessive or totally absent and unhelpful, especially for DMs. I have a great deal of interest in producing a D&D where casters do not rule the roost, where it is a teamwork experience and every player, no matter what aesthetic preferences they may bring or interests they may have, can find a character concept that actually equips them with fun, exciting, potent, personal tools to contribute to the group's success, whatever form those tools may take. I have a great deal of interest in giving diverse, flavorful options, such as races/ancestries and feats and weapons and spells, that are sufficiently close quantitatively that the players must instead make qualitative decisions about what they wish to do, because that actually makes for interesting choices, not dull calculations.
Your end goals (italicized above) are laudable. How do we get there?
In my view, your suggested means in fact work directly against the very goals you seek to achieve. Why is that? Several reasons:
--- by far the best way of designing teamwork into the game is for classes (and-or species?) to have hard-drawn niches - things that only they can do the least bit well (though anyone else can always try, their odds of success would be very low). This by definition somewhat limits the number of classes and-or species, as there's only so many viable niches* to go around: healer, front-liner, archer, scout (outdoors), scout (indoors), blaster, burglar, charmer, diplomat, investigator (a.k.a info-gatherer), supporter (a.k.a. buff-provider), and maybe tactician. Front-liner can be subdivided into defensive front-liner, attacking front-liner, swachbuclker, and a few others maybe; and sub-niches e.g. undead specialist or artificer might also have a place. And I probably missed a few other obvious ones.
Then, give each class free rein over one of those niches and extremely limited access to the rest. Nearly all abilities come baked-in with the class as it advances in level - which also neatly makes the game much easier to learn as there's fewer choice points and options. Fact of life: the more choice points there are and the more options there are to choose from at each one, the harder the game is to learn. The intent here is that every character is or becomes a specialist at what it is good at (i.e. its class), and making that believable in the fiction is trivially easy.
Oh, and make multiclassing possible but intentionally noticeably sub-optimal, to encourage playing single-class characters.
And yes, this very intentionally leads to "If you're a Rogue it means you can do this this and this but can't do that that and that", because that's the whole point. That's where the teamwork piece comes from: no character can do everything on its own, you need support from others to cover your weaknesses.
Along with this, a return to clear and obvious mechanical differences (both beneficial and penalizing) between the species, including baked-in stat bonuses and penalties (except Humans, who are the baseline) would also help; both to de-blandify the game and to make species a more meaningful choice. So, in the end, choosing species and class would be the biggest choices a player ever has to make, after which everything else falls into place; and a DM can always step a new player through that choosing process during roll-up.
* - you might notice "controller" isn't listed as one of the niches, and that's very intentional: controllers tend to instantly become the most powerful class(es) and thus ideally that niche would somehow be either downplayed or spread out between as many classes as possible.
And the fact that even a single request for "hey, maybe usability, excitement, or balance could be a consideration that sometimes is just a little bit more important than absolute and unyielding fealty to 'verisimilitude' über alles" gets met with "oh, so you want the gamw to be bland, generic, flavorless mush that means nothing, has no story, and turns everything into samey garbage, well not on MY watch buddy!" is EXACTLY why I find this so infuriating.
You literally cannot even criticize excessive commitment to "verisimilitude" in the smallest degree without being accused of wanting to destroy the game.
You can have verisilitude and-or believability at the same time you can have usability and balance (excitement had to come from the player, not the design); though I'd put usability ahead of balance were I forced to choose.