"Real" depreciation is reflected in individual houses...and that is one reason adjustments for "age" "condition" etc. are inprecise at best in the Sales approach. But if we use a book to calculate SF costs, then we can use their depreciation (residuals) charts in lieu of straight line or we can modify the depreciation to account for "short term"..(functional) and called it Modified age-life. Likewise, USPAP counsels us to be aware of and apply those methods, techniques, etc etc that our peers do and what the literature teaches. Applying our own standards or biases simply has its limits. And if you don't have at least one basic text covering basic principles...get one.
All of it is a "guess" of sorts...supported, but certainly not rigorous scientific proofs. I mean how many of us have appraised a property and claimed it has 20 years remaining life to only go back 5 years later and it is gone or abandoned?...or like the other thread about the $42 mil. "tear down"...
However, the difference between the Land value + Replacement Cost less physical depreciation should it differ from the sales price is a reflection of the obsolescences (as opposed to the deterioration). And that measurement when applied to the comps, should give you an understanding of where the actual CA MARKET VALUE is...and typically, if you have done it right, then the CA and SA should not be much different and that difference can be attributed to the precision of the measurements and the efficiency of the market..or lack of efficiency (volatility, risk, whatever)...