Good point, and rubbish-n-roadmaps is very apt (and google says it's original.)

But I did cover myself on this: if you do solve an unsolved problem, all is forgiven. Getting it out into a closed beta can tell you about a large miss earlier, with less investment, and less reputation risk. Of course, it would always be better to do everything right in the first place, if you can afford it. Rubbish-n-roadmaps is the worst approach except for the others.