All the more reason.

Getting deep insight from half a dozen key people takes no more effort than many other kinds of "basic research". IMHO, it's an excellent expenditure of resources. Good feedback -> all the more reason to go gung-ho, bad feedback -> save yourself a lot of grief later.

How many excellent products never saw the light of day because of lack of demand? I'd do everything I could to prevent that from happening.

That's why it's super important that you can hack up a prototype very quickly. It may help to learn a RAD tool, or Lisp.

Absolutely. I consider "hacking up a prototype" part of analysis, not development.