My problem with Fannie's approach on this is that our professional standards don't use the term "accurate" with respect to opinions and conclusions. By definition, accurate connotes a degree of finality and precision that appraisers generally avoid. The operative term we use is "reasonable".
My other beef with it is that criticizing someone else's opinion is hard to do. Because a value conclusion is an individual's opinion, you almost have to get inside their head and prove an element of intent.
I've done enough reviews to know that an appraiser basically can't come up with an unreasonable value conclusion without distorting some of the facts. It's true that there is some subjectivity with respect to categorizing quality and condition and "most similar" and such, but it's usually not enough to result in big variances between appraisers.
As a reviewer, I look for the big lie. If the subject description checks out and the facts of the sales transactions check out, the only other question that remains is whether those sales can reasonably be considered to be among the most recent and similar sales, as is certified on the Fannie report forms. If they are, and absent any really unreasonable adjustments or some dumb technical error, then the results are going to be reasonable overall. Or, at least not so unreasonable as to warrant a complaint.
So no, I don't think value alone should be a valid basis for a complaint. If a reviewer can't demonstrate how that value conclusion crossed over into unreasonable all we're left with is the "I know a bad appraisal when I see one" gambit. That's a mode of appraisal review that I categorically reject.