The problem is appraising methodology was developed before the mass data gathering of computers and ability of anyone to access data over a smartphone. Now, everyone with access to data is an "expert" , qualified in their minds to challenge an appraisal.
Add in after delivery of an appraisal clients or Fannie , reviewers at an AMC, use computer programs and software programs to vet results that a human came up with using human qualities of judgement, analysis, opinions. Appraisers are not just analyzing data, we are analyzing markets and buyer/seller behavior and factoring in market trends to the data. Which does not always result in computer friendly consistency. If appraisals start to mimic computers to deliver computer friendly results, such as using AVM type programs , then it is no longer an appraisal. Call it something else because it is at that point something else.
CU is a programmer's ideal that comparing appraisals to each other will reveal which is "right", or "better" or whatever they expect to find.
Except for outright bad/misleading results such as calling a C4 house C1, differences between reports are intrinsic to each reports' development.
Appraisal methodology was developed to be challenged by review performed by another appraiser (field or desk review). The methodology was not developed to be "reviewed" by computers, software programs, data dumps. or agenda driven ROV by non appraisers. Which is why appraisers have so much trouble coping with and responding to these kinds of challenges.
How do you respond (without crying or tearing hair out) to a set of "comps" clearly superior and non relevant sent as an ROV? How do you respond to inane set of non relevant "comps" spit back from a computer? How do you "explain in detail" how an adjustment was derived, when the method of developing that adjustment is partly intuitive aka judgments and conclusions of filtering data vs raw data math. How to respond to a CU blind type of question such as your adjustment or condition rating differs from a peers' or model? We respond as best we can but it's difficult to form defensible responses to a series of challenges that appraisals were not designed to withstand.