That leaves us with the question: Does the domain knowledge accumulated in a PhD program lead one to later find problems that they would have never otherwise been determined to solve?
My hunch says it depends. Some people will find a solution and then go looking for a problem. Others will find a problem, and then go looking for solution. For the former group, a PhD will be quite valuable, but not so much for the latter group – the chances of them even having studied the right domain in advance is slim.