This "but the energy cost is net negative against keeping what we have" is an odd basis of reasoning. Some days I think its just a modern variant of "we should never have come down out of the trees" thinking.
No development can get away from a positive energy burden against the sunk cost of history. For active (coal burning power) vs passive (solar) the net energy cost is probably always good now: the cost of deploying solar has dropped enough that the sunk cost of the coal burner is now exceeded by the future lifetime upside of the PV, even given its carbon debt to make.
You're in a niche, arguing that scavenging existing clay tiles and keeping what you have is net better carbon output (bearing in mind your power has to be generated, and transmitted) than replacing the tiles with solar tiles. Thats different to 'replace coal burner with solar' but it can be calculated.
TL;DR you're asserting you think its worserer. I think its an interesting question, I don't actually think it is, but if it was, the question doesn't terminate in the sunk cost of the existing tiles, its the sunk cost AND FUTURE BURDEN of the coal burner you don't use. It's not just the production cost burden of the solar PV tiles, you have to remember to include the CO burden from the energy source you don't displace by making it locally.