My anecdotal evidence regarding CSS selector performance is: I’ve never once bothered with optimizing for it, and typically outright ignore traditional “advice” around it, and it’s never been a problem.
Ben Frain looks around at what browser people say about it and:
At this point I’m happily re-concluding that CSS selectors are rarely a problem with static pages. Plus, attempting to second guess which selector will perform well is probably futile.
If you’re doing something extraordinarily DOM-heavy, I think you know that and can worry about it then.