In which cases would you most prefer such scans be made?
b. I expect that in that case we'd use it primarily to inform sentencing and parole decisions, and that might actually improve those decisions. (I'm not particularly confident that it will, though... say 30%?)
I'd actively prefer we avoid scanning in cases d and a, except perhaps as research on populations, as I don't trust us to do anything even remotely sensible with the data.
I'm not sure about c... it depends mostly on whether we have any way of intervening usefully in their subsequent development.
I have just received a survey questionnaire regarding future directions in EU (European Union) research funding, and thought it would be interesting to see how LessWrong would answer the main question:
Imagine that EU funding is available for one ambitious, visionary project extending beyond 2020.