“When cost drops low enough, any surface with any exposure to sunlight is in scope for installing solar on if it can yield more energy than the cost of installing solar on it.”

I don’t see the logic? As panel prices drop, installation costs start dominating overall cost of energy produced and so economic pressures will be on simpler and simpler PV installations. Laying a panel almost flat on the ground amongst thousands of others will be a cheaper process - both initially and in terms of ongoing maintenance - than anything that involves sending workers up ladders or cherrypickers attaching and wiring a panel onto the side of a building.

So grid scale will start to dominate over domestic or “novelty” (e.g. floating panels on reservoirs) and simpler and simpler approaches to installation (like removing tracking) will become attractive.

These effects are already apparent. Basically the opposite of what you’re predicting?