And there's a stereotype that a player who has a character with a lot of abilities from expansion books is going to be much more powerful. Why is that? The expansion book choices aren't inherently more powerful. But firstly, because of normal variation, some of them will be more powerful than average and some less so, a player looking to make a character effective will be drawn to the powerful one. And secondly, there'll be something that has exactly the right combination of abilities that are worth the most to this character, a lot more than they would be worth to an average character.
This is also what I observe about the human variant race. In theory "+1 to any two ability scores" might not be worth more than "+1 to str and cha". But it IS, because there's only so many races in the core book, and they won't necessarily be a race that excels in the abilities that a particular character benefits from the most. So "+1 to any two" is worth more than "+1 to these particular two", because it lets you mix and match the parts worth the most to you.
I struggled for a while with how to cope with this with players with a range of optimisation options. If all the players play "My character uses a dagger because it's in character, I know a sword would be better in every way" or all the players play "not the MOST powerful character, but I took all the obvious improvements I could" then it's fine, but how do I cope when there's some of each?
If the game's not combat focused it doesn't matter, but if it is, it's usually not satisfying to have some characters be just better than others. I eventually realised, I could mentally chart HOW optimised characters were, on a scale from "just picked options that sounded in character with no care for effectiveness at all" to "picked options that make sense together, but didn't try to optimise" to "made the obvious choices of choosing the most useful weapon and armour available and focusing on the stats most useful to my character" to "seeking out specific options which work well with my build". And I could recognise, given the player preferences, where on the scale made the most sense. And guide people to that point, either by giving them bonuses or suggesting tweaks, if they were less optimised. And suggesting they deliberately forgo some more effective choices in exchange for some more fun but less effective ones if they were more optimised than that. And that *should* work.