But some societies do that way better than others, so it would seem that the appropriate place to intervene is at this level.

If this is not the appropriate place to intervene, then the whole idea of multiculturalism seems exceedingly ill-advised. Now, I do admit that multiculturalism has encouraged some non-negligible problems of the "familiarity breeds contempt"-type, so if you want to press the point, I would concede that the jury is still out to some extent. I'm hopeful but can't offer incontrovertible evidence.

The key idea behind multiculturalism is not a lack of presumptions, but rather tolerance of outliers. We don't give genetic testing to kids so that we can deliver poor nutrition to the naturally tall and superlative nutrition plus growth factor enhancement to the short so we all end up the same height, nor do we provide much support for people wishing to do the best they can to surgically/medically alter their height--even though height-based preferences are a real thing, with measurable discrepancies in income and so on that often approach that seen with racism (at least in the U.S., U.K., etc.). Rather, we keep trying to impress upon people that, you know, one's worth as an individual shouldn't depend on one's height; people are all people.

But there's still no problem with expecting that tall people are more likely to play basketball. If they don't, no worries! But we facilitate that path because it is one that works well.

Now, one might counter that the sex=gender path doesn't work well. And my reply to that is: well, maybe not, but how do we know? The conflation is very common among human societies, even among those that have multiple genders. Is it entirely a problem wherever it appears, just to different degrees, or is the facilitation of the normal path actually important but in our society we have little direct experience with how badly things can go wrong when it's not facilitated because we have not seriously tried the other way?

Regarding the "kinda lost"ness, I don't think we're asking the right questions to even know the answer, in large part because of attitudes like that of the original commenter's.

For instance, suicide attempts among young people who identify as trans are somewhat lower in more supportive states. But the suicide attempt rate is still shockingly high, and there are way more people who identify as trans in more supportive states (see, for instance, https://www.nytimes.com/2022/06/10/science/transgender-teenagers-national-survey.html).

This is not what one would expect if the reaching out to "kinda lost" people worked only in the way you suggested. Now, people are complicated, and societies more so, but anyone who doesn't look at that and go, "Hey wait, does that mean that on balance, trans-positivity makes people want to kill themselves? Hadn't we better figure out whether we're overall causing more harm than good, and if so, see if there's a way to switch it to more good than harm?" isn't actually acting out of genuine concern for the human condition, but rather out of whatever self-interested, ideological, or proxy-goals they've set for themselves.

In rejecting destructive hatefulness--which we should absolutely do--we have to be careful that we do not polarize ourselves to the point of blindness in opposition, and thereby go careening around causing our own perhaps slightly less spectacular destruction and harm.