That's likely because those who define and promote it work very hard to conceal the fact that there are many many "ask-the-human" type decisions and factors involved. No two balls hit are "the same" in nature; they are only "the same" when so perceived by a human eye and significant features (e.g., 'speed') defined as relevant by a human brain. The same goes for the definitions of atmospheric conditions, etc., and any number of refinements that could be build into these models. There are some types of 'neural network' types of analysis (insofar as I understand these!) that I believe can eliminate this 'ask-the-human' step, but UZR (at least from what I can tell) is not one of them. Mind you, I'm not opposed to this type of analysis, and it certainly beats the kind of approach critiqued and ridiculed so well in Moneyball (the scout's 'feel' for the game in particular!). But then, I'm not opposed to judging hitters by batting average either (or OBP or OPS). Statistics help. But few methods I've seen are completely objective (maybe the one my adolescent brother once used for a whole season: all he recorded over the year was a ratio for each AB: number of bases achieved vs. number possible. So 1-4 for a walk with no one on; 3-6 when a weak ground ball goes through the SS legs allowing the runner on second to score. No one, of course, either a traditionalist or a sabermetrician, would accept such a system as a standard on which to judge others).