I feel like the whole genre of 'How to tell if someone will perform well' is trying to squeeze blood from stone. I've worked with a lot of different people, and the more broad my experiences become, the less I think I am able to tell who is good and who isn't.

I've worked with people who have extraordinarily high output, but slowed everyone around them down because they filled absolutely no-one in on their decisions and changes, so we were all left trying to beat the next code-base warping idea.

I've worked with people that accomplished almost nothing, but lifted morale so much that they more than made up for their lack of productivity. (I wouldn't recommend looking for this scenario though, most low-output individuals drag morale down...)

How many marriages do you think would make it 5 years down the road with only a month of knowing each other? A week? A handful of 1-hour interviews? I mean, that's the kind of time commitment you are looking at with a job.

Just filter out people who are obviously bad fits, and fire fast. The diminishing returns of trying X, Y, and Z, to me, just aren't worth it, and if you are a small company, you've got better things to dedicate brain power to.