Permanence is priced in already and I think sellers know it

Stripping the benefit out entirely as a first pass is the standard discipline, and the reason is exactly the compounding you described. Substantial improvement is measured against the basis of the building, and the land allocation sits outside that test in most structures, which gives sponsors an incentive to push allocation toward land to shrink the required improvement spend. That allocation has to be supportable, and how much room you have depends on an appraisal that will survive review. A tax professional should sign off on any allocation you're relying on, because an aggressive split that gets recharacterized undoes the compliance position rather than just costing you money.

On the seller pricing, what you're describing is the benefit leaking to the land owner, which is a known feature of any place-based incentive. The counter isn't better negotiation on the same asset. It's sourcing where the buyer pool doesn't know the tract status yet, which after the 2027 designations take effect means the window is widest right when new maps publish and narrowest eighteen months later.

What I'd add to your math: run the deal at your unlevered yield on cost with zero tax value, then run it again assuming you exit in year eleven rather than year ten. The ten year hold is a floor, and real exits slip. A deal that pencils on a year ten sale and turns marginal on a year twelve sale is a deal where the tax benefit is doing the work you claim you've stripped out, because you've embedded an exit date you don't control. Also check whether the enhanced rural provisions apply to anything in your pipeline, since the improvement threshold treatment there differs and it changes the required spend on qualifying rural assets. Confirm the current rules in writing with your advisor before you underwrite to them.