Every gnome isn't inherently a better wizard than the greatest human wizard. But exceptionally great gnomish wizards have the potential to be greater wizards than exceptional humans, as long as various divine entities and over the top artifacts stay out of the picture. Which they usually don't. Likewise, every half-orc isn't going to be stronger than every halfling, but an exceptionally strong half-orc is reasonably going to be stronger than an exceptionally strong halfling. How is that remotely controversial?
Also, I seem to remember that Tasha's stat change rule was optional. And all it does is let people power game while pretending to RP, even though that power gaming has absolutely nothing to do with RP. To elaborate, please do try to make a non-silly explanation of why your paladin needs a dip (not a full conversion, just a dip) into hexblade. If you're remotely honest with yourself, you know it is total nonsense. Your DM knows it. The people you're playing with know it. And you getting to do that just because it is powerful and meta sets a bad example for everyone.
Likewise when you want your exceptionally strong halfling to be just as strong as an exceptionally strong half-orc. It is not good role-play. And your reason for wanting it is terrible. You don't have any real justification for how it makes sense, you just really want a bigger stat bonus for your class so you can power game better.
This is a good and sensible take overall. I'd like to counter two points though.
I would love to play a system where the strongest half-orc has the potential to be stronger than the most exceptional human or halfling. But DnD5e is not that system. It never was. The upper cap is 20 and everyone can reach it. Strength is an abstraction of one's physical strength only in part - it's also the abstraction of one's ability to strike with a weapon (proficiency only pertains to experience with specific weapon groups). Githyanki get +2 to Strength despite their thin arms because they generally all undergo military training. I'd have much more respect for ability bonuses if they raised the upper cap, but they don't. They're arbitrary.
The second thing I'd like to counter is that the half-orc vs halfling strength comparison is an extreme example, not to mention a halfling can already be as strong as a half-orc anyway. What most people arguing for removing the bonuses want, I believe, is to be able to play a smart halfling wizard or a wise orc druid without playing second fiddle to the human. Starting with a decent number is not min-maxing.