Steven-
My suggested statement is "my" general statement; in my reviews, I then list the specifics that I think are significant. As a reviewer, I find inconsistencies or errors to be in two classes: Trivial and Non-Trivial:
The trivial ones could be typo's, the wrong specific zoning description (as long as the zoning itself is correct and the subject's conformity or lack of is correctly indicated), perhaps a date on the comp (1996 vs. 2006, when everything else matches with my 2006 data), etc. A trivial item or even a number of trivial items may have no impact on the overall credibility of the report. Some, while trivial to me, may be significant to the lender; the appraiser's correct license expiration date, for example, when I've already verified the license is current and valid. As a rule, these are not "significant", although I may list them for my client's benefit. When I do, I state that I do not consider these items "significant".
If enough trivial issues are present, I may begin to question the quality of the report. A sloppily written report detracts from its quality and credibility, which I am charged with forming an opinion about. If this is the case, I do consider this significant, and again, state my reasons. However, the value may still be supported in the market. This is an example of a report that "hit" a reasonable value but is so poorly presented, one has to question the competence of its reporting, if not development process. We read about these instances all the time.
A non-trivial issue is one that I find significant to the credibility and usually value of the original report. These can be when comps are provided, a general "these are the best comps" statement included, yet I find what appear to be more proximate, more recent, and/or more similar sales. Or, when a subject is a 2br home, all comps exceed 2br in configuration, and there is no explanation about market acceptance of 2br homes or the appropriate market adjustment between the two (the buyer pool for a 2br home is typically different, and smaller in my market than that for the 3br home).
It is when there is a conflict between what is in the report and what I discover in the market, and that conflict leads me to analyze and conclude a different opinion of value that I state the report has inadequately explained its adjustment and/or comp selection. And, again, when I make that statement, I'll provide the evidence I'm using to support my statement.
Adjustments for GLA are not precise. Rarely do I question the amount. I sometimes question it if the report's application isn't consistent in its results (and the difference has to be significant, not "trivial").
You in your assignments may, as a practice, use more than the 3 or 4 comps that are typically found in most residential mortgage finance appraisals. Your analysis is probably a little "deeper" than what is found (or required, given the intended use) of the residential reports I typically review.
3 or 4 comps can be more than sufficient to adequately support a value.
It isn't the number of comps that is necessarily important, it is the comps used vs. what appears available in the market that becomes a non-trivial issue. And, when there is inadequate comment or analysis provided so that I, the reader, cannot reasonable understand why the report has one set of comps or uses one set of adjustments when the market data I'm looking at (presumably, the same available to the originator) indicates another set appears more appropriate, then it is accurate to say the report has inadequately explained itself. I did use the terms "Analysis/Comp Selection", and those terms may be interpreted differently. But as a rule, my client (the intended user) knows what I'm talking about and so does the originator if provided a copy.
As far as the boilerplate comments, as a rule, they don't bother me, and I have some myself in my reports. But, if the boilerplate, which is typically a "generalized statement" is used as a substitute to explain an analysis process that I can't understand, support in the market using my own research, and certainly can't replicate, then it is inadequate.