Cost related adjustments are common in many reports I've seen over the years. I don't see them as particularly accurate, but I know appraisers yet that apply depreciation to fireplace adjustments by using the M & S "cost" of a fireplace...pool...outbuilding. OTOH...
When I try to extract a fireplace in a regression, I may get a reasonable answer but normally it gets either a ridiculous answer or the R square goes to pot. Boot it out and the R square is much higher. Or in this example, it was pretty reliable - ended up $5,480 for an adjustment for fireplace. I don't see a need for using cost for a fireplace, brick v siding (kind of a quality issue) garages, etc.
View attachment 59307