Ultimately, what it mostly comes down to is whether certain skills have been cultivated in you from a young age or not.
This is very true, but often unmentioned - it isn't as helpful for the purpose of preparation (not that preparation in general would be very helpful for UMAT), and isn't very reassuring (it'd be like saying "So, how'd you go in that test-where-you-have-to-have-been-raised-a-certain-way-to-do-well?"). Also I'd hazard a guess that people who do get good marks prefer not to attribute it to the thing they did growing up, to avoid sounding arrogant, if nothing else.
Although I guess there are two very important messages in there for health sci's:
1. UMAT is not something we can prepare for, not now, anyway, so don't stress too much about it.
2. UMAT is essentially not under our control (yes this sounds bad for confidence, knowing it's not under control, but still), but HSFY results, a far more important factor in determining admission to med, are under our control, so it's much more important to focus on doing well in HSFY than to worry about UMAT.
EDITed to include the content below:
Whether or not the skills it is testing are applicable in medicine is another story. For example, I can't imagine anyone asking their doctor to pick which picture comes next in the series (section 3)...
Actually, that's why Section 3 is only worth 10% in Otago's book. Section 1, where we deduce something based on information we've just been given (not unlike diagnostics as far as I'm aware), and Section 2, where we work with the messy business of human emotion, seem to be at least on the right track towards being relevant (whether it actually predicts these qualities in doctors AFTER they've been trained, though, is another matter.)