If you read through 38 CFR you'll find a lot of rating processes are purely subjective and leave it open to the rater to make determinations that are technically beyond what one would expect a layperson to be capable of doing.
In fact, the wording the Dr's who do the evaluations are required to use leaves a lot open for interpretation. Terms such as "as likely as not". Very ambiguous, if you ask me. I'm more of a concrete term kind of person, in other words, it either "is" or "is not".
I think if everyone in the system knows and understands the terms this is not a problem. But that is a big "if." Worse for the veteran, because they are not trained in this field. Unfortunately, after their service they are thrust into this complicated system where they must either educate themselves or trust the government. Neither choice seems fair to me.
I find the disparities disturbing. On its face, such disparities based on geography show that there are many supposed experts in the system who do not understand the terms nor their application. And unfortunately, it appears that having an attorney to advocate is neccesary to get a fair shake.
Even more disturbing to me was some of the testimony I recently read from Mr. Daniel Cooper, Under Secretary for Benefits, DVA, who stated in his testimony before the Senate, (http://veterans.senate.gov/public/index.cfm?pageid=16&release_id=10805&sub_release_id=11054&view=all) that is takes two to three years for new hires in the adjudication process to become fully productive. It makes me wonder what happens to the veteran whose case ends up being decided by a "new hire." I also was concerned by the testimony of Mr. James Terry, Chairman of the Board of Veterans' Appeals. ( http://veterans.senate.gov/public/index.cfm?pageid=16&release_id=10805&sub_release_id=11057&view=all ). Mr. Terry testified that the remand rate at the BVA (the BVA sending the case back) is currently 32 percent. I read that as meaning that there was not enough information developed at the RO level or below to correctly rate a case. That means that there are serious problems in almost a third of all cases. At the BVA, Mr. Terry testified, the error rate (defined as reversal or overturn by the Court of Appeals for Veterans Claims) is currently 7 percent (which I question because all cases challenged have not been decided yet at CAVC, and given the rate of remand or awarding of higher benefits discussed in GAO report, below). That is astonishing to me given the fact that nearly 60 percent of cases at the CAVC are filed by veterans without the assistance of an attorney. Historically, the system has failed at an alarming rate. As late as 2002, the Government Accountability Office reports that half of cases at CAVC result in granting of higher benefits or remand (http://www.gao.gov/new.items/d02806.pdf). To me, this suggests a near catastrophic failure in properly rating veterans.