so then what are they buying in this segmented model sale???
It's a model....so is the scenario laid out. He wants to know how to arrive at a contributory value to the LAND value...or is there any.
Your home is taxed without selling isn't it?
You wouldn't do a cost approach on an old farm house, old barn and workshop.
Actually I do a cost approach on every large tract rural property. ALWAYS. First I want to determine the value of the land as if vacant and available for its highest and best use. Then extracting the value of improvements from other sales - starting with subtracting the land value to get the building residual then applying similar accrued depreciation. Deconstructing every sale is necessary to determine if and how much obsolescence is present.
So first I want to know what the LAND is worth. And the larger the tract, typically I find that the more depreciation accrues to the buildings and house. And my assumptions start with the idea that the oldest improvements (non-house) are the least valuable.
So say my sale is $476,000. The land is worth $400,000. Therefore the net to improvements is $76,000. So say there is a house and a barn. The house we can estimate the $/SF from similar sales or even straight or modified depreciation. Let's say that it is $66,000 you allocate to the barn and site improvements. From other similar sales you have determined a typical site improvements (well,septic, gravel drive) value is $6,000. That leaves $4,000 for the barn. If the barn is only 2 years old and cost $20,000 to build, then it has lost most of its value, and has depreciated by 80%...far in excess to its physical wear and tear. It suffers functional obsolescence in the market.
That is why you always do the land "as if vacant" first. Since no 2 appraisers are likely to extract in exactly the same way (except all should be extracting the land value as if available for HBU) the appraiser must perform the same analysis for subject and comps consistently.
And to answer the original question, both sales and cost approaches should yield similar results assuming you have sales data to develop a good depreciation for the improvements. And I find too many appraisers apply depreciation rather mechanically and cannot understand why the cost approach comes in "high" to the sales approach.