I actually think standardized tests are a very valuable and necessary part of any admissions process. But what I resent about the MCAT, more than with any other standardized test like the SAT or LSAT, is the enormous potential for raising one's scores in a relatively short time with a $2000 prep course or other intensive prep strategies focused on cramming in content and test-taking strategies. This puts prospective applicants on uneven ground based on access and ability to pay for these options.
The math/verbal SAT's and the LSAT test critical thinking, comprehension, logic skills and, dare I say, to some extent plain old basic intelligence. You cannot buy that. Of course there are test taking strategies but there's really no way to "teach" someone who just doesn't get it and hasn't already mastered the material over years and years of formal education. I personally think they are a better measure of aptitude (not necessarily a great one, of course, but better in comparison).
The MCAT on the other hand (aside from the verbal section) focuses far more heavily on content knowledge which can be crammed and regurgitated by probably just about anyone given enough repetition and practice. Yes, even if you know all the science information but lack critical thinking skills you won't score super high. But someone who is highly intelligent and capable of learning very complicated material, and critically applying that material later on in practice, just doesn't have a chance of scoring very highly if they don't know certain specific definitions, reactions, formulas, etc. Which is where paying outrageous sums of money for intensive prep courses makes the difference- unfairly IMO.
You read stories of people here on SDN who took all the pre-req courses and then got like a 25 or 26 on the MCAT. Then they went and took a Kaplan course and two or three months later they got a 33. I guarantee you wouldn't see variability like that following an SAT or LSAT prep, no matter how expensive. So how valid and reliable a measure of aptitude for medical school can it really be if the scores can vary so much in a such a short time?