Damn that's super interesting. The test itself is useless, but self-identifying into a the scheme accompanied with the aforementioned useless test helps you find likeminded people. People you allegedly find that you work better with. It's a shame that there is no way to self-identify into anything else, and find like-minded people, without relying on a thoroughly antiquated pseudoscientific bullshit classification system. My point is this: Praising MB for letting you self-categorise into a scheme that, on the surface, generates better team dynamics than no system at all is giving credit to systematising. It is not giving credit to MB. Try this self-categorisation system out, and see if it out-performs no system: knowledgeable / unknowledgeable.

Depends on what you're trying to achieve. If you're looking for insights on yourself: then being asked a question, to have the answer you gave fed back to you with a pretense of insightfulness will easily be outcompeted by staring at a mirror for 10 minutes. If you're trying to generate complementary group dynamics: sort people by interests and spread them out.

If you're trying to generate efficient groups: sort people by interests and don't spread them out. Literally the worst kind of systematising will go toe-to-toe with MB.

There isn't a better alternative because the premise you can understand and predict behavior by assigning a few labels based on a few specific questions is bunk science