Yea true, if people can vote on something, other people will use those votes as metrics for how good something is

My perspective was more about what they actually do. Not the meta-effects they might have socially

Eventually, you will be able to turn a repository with a high star count into money or advancement

I think you overestimate how much money or advancements you can really get from it though.

Money wise - I can't find an overview of "Most Sponsored github repos" - but it's pretty bare. I checked to see if I could find any example, for example if you look at FluentAssertions - A project that basically everyone uses, has 292.6 Million total downloads on Nuget. If you check their sponsers - they currently have 17. Assuming their the lowest tier, you're getting $85 a month. Which is cool, I guess, but a neglectable amount for a developer with a normal job

And advancements wise - any actually good developer doesn't really have a problem getting a good job - And any good company reviewing a candidate might fool the HR by buying stars, but a dev reviewer or something will actually look though the code won't care much about stars

I don't think looking at the star counts makes you automatically a bad developer, but it certainly shouldn't be the only thing you look at. If you're unfamiliar with libraries solving a specific problem, I don't see anything wrong with looking at them from the most to the least popular. Popularity can also be a sign of community and therefore more likely continued "support"