So you state that there are at least two axes -- severity of the problem and how widespread the problem is -- that influence whether a problem is worth addressing (or whether it is a "real" problem, as you put it). I agree so far.

For instance, we can probably agree that slavery was an issue that, in the US, affected a small minority of the population, but that was extremely severe so worth addressing and a "real" problem.

In the same vein, I would state that police brutality towards black people (typically categorized as a "woke" topic) in the US is a problem that affects a minority but that is very severe. Not as severe as slavery, but severe enough to be a "real" problem worth addressing.

Not using preferred pronouns is perhaps another rung down on the severity ladder compared to police brutality, and affects fewer people. But I think it's still a topic worth talking about because it scores very highly on a third axis we've omitted so far: how easy it is to address the problem. Not using preferred pronouns is done exclusively out of callousness. It is similar to other forms of bullying: it only affects a fairly small subset of people and the damage is "only" psychological, but the solution is: just don't bully/misgender people.