You make some good points, but when there are cheaper alternatives, engineering problems ARE showstoppers, and for good reason! If Musk announced that he wanted to build a datacenter inside an active volcano, my response would be pretty much the same - "Why?"