IBM has lots of experience cooling chips inside mainframe computers. I saw a grid of 192 processors each with a metal rod in contact that conducted heat to a coolant fluid. This has long been the limiting factor in concentrated photovotaic conversion.
The intermiitancy of sun light problem remains.
Plus what controlling entity is going to allow ugly parabolics on the landscape?
In my opinion, solar has the potential to be effective at shaving off the electrical peaks associated with cooling during the heat of the day. As a base load it is quite difficult without significant advances in storage cost/efficiency.
I HOPE that there is a “scram” like system to move a counter mirror into the beam when main “core” coolant fails else the cost of a simple water leak is going to be pretty high.