Relevant, perhaps. But essential? Not really. As I've said, if the skills tested by the UKCAT were so indispensable at the point of entry to medical school, no medical school that didn't require it, or something similar to it, would get GMC accreditation as they do at present. Moreover, it would have been introduced sooner than 2006. It's been shown to be only a mild predictor of success once at medical school and is dwarfed in that respect by A Level performance.
There also doesn't seem to be any noticeable gap in the performance of junior doctors produced by Bristol, the only medical school not that doesn't require the UKCAT or any other admissions test for its medical degrees for 2016 entry, and medical schools that use the UKCAT. So yeah, I see it as little more than a pointless test whose primary (and, I would argue, only) purpose is to provide another discriminator for admissions tutors.
Interestingly, the last two medical schools to adopt it, Liverpool and Birmingham (using it for the first time this year), seem to have done so solely because of the pressure placed on their admissions tutors by a very large number of applications from people who didn't do well enough on this one test to stand a chance elsewhere. Bristol, too, rumour has it, is going to introduce the UKCAT for 2017 entry for precisely this reason. Are we to believe that they suddenly realised that their applicants lack some crucial skill?
I don't, anyway.
I can't speak for the GAMSAT because I've never done it or even looked at a practice paper but, although I liked the BMAT a lot more than the UKCAT, I have similar feelings towards the BMAT, yes. I have strong views about education in general which are rather too long to go into here, but the long and short of it is this: I don't like tests/exams that are difficult because of the timing rather than the content. The idea of having to guess a question that you could otherwise do (as I had to do in the BMAT) and move on is abhorrent to me, as is the idea that exam technique should be more important than what it actually tests.