What's idiotic is to criticize any particular movie for "failing" the Bechdel test. But as far as I'm aware, that's not usually how it's used; it's a tool for statistical analysis of groups of movies, e.g. all movies from a certain country, or all movies released in a certain year. Maybe I'm wrong on that, though.

The goal can't and shouldn't be that every single movie meets the criteria of the Bechdel test - it's easy enough to think about existing or potential movies in which there isn't no or very little room for female characters and adding them wouldn't improve the movie in any way (movies set in male boarding schools or prisons, for instance, or war movies set in historical eras). But looking at which percentage of movies, or which percentage of movies of a certain category, fail or pass the Bechdel test, and how that percentage evolves in time, that's not so stupid. And that does say something about gender bias.

For an individual movie, the Bechdel test is at best a rough indication, but it will indeed have lots of misses in both directions.

That said, I'm not sure if it's as easy as you suggest to find an alternative, equally simple test for gender bias, even for evaluating individual movies. Of course you can have far better methods to analyse movies which are less likely to miss in either direction, but those aren't as simple to define or execute. The only modification that comes to mind is that perhaps the test shouldn't make the presence of female characters a requirement, but should rather say, if there are female characters, then do they have conversations that makes clear they go beyond being romantic interests. If there are no female characters at all, that's unlikely to be a sign of gender bias; more than likely it's a direct consequence of the chosen subject of the movie.