For physics? Probably not. But physics grads don't make up a particularly big part of the economy; it's actually not an especially great degree to get if your primary concern is a STEM-y career.

For HR directors, purchasing managers, corporate controllers, practice managers, financial advisors, and jobs like that? Yes, I think you can devise filters for anything colleges filter for that would be both more effective and far cheaper than college.

> Can you devise a filter that's as rigorous as all the tests and projects that a student would have successfully completed over the X years it took to earn the degree?

Eventually I don't see why not but I don't think it's an easy thing to do nor do I know how to do it for a physics R&D type of position. Obviously education isn't the only criteria people use today anyway (I can't tell you how many CS grads I've interviewed who really struggled or flat out didn't know a lot about CS).

It's an age-old problem: how do tell if someone can do what you need them to do without wasting your time and money? As the skill and knowledge requirements go up this becomes harder and harder to do, with or without the degree shortcut.

For our profession? Yes. For example your github repos, SO profile, the app you've built or your projecteuler stats may prove better filters.

But I don't need to - smarter people already did the work for me. This approach is already practiced (mini scale I believe) in our industry.