John,
As I say to my boss, he pays me to have opinions They might be ill informed at times (though I try not to), but I'll have them. So I'll be more than happy to contribute where I can on the hope it'll be useful. We all have a vested interest in seeing Gen succeed, so while we may differ in opinions, we always have a long term goal in mind.
Having said that, members of the community will be inherently selfish in their ideas (which includes me, and there's absolutely nothing wrong with that!). They have their own needs and goals, and will generally not need to worry about the longer term picture for Gen. They'll be looking for satisfaction on a shorter time scale than you will be, and may see no immediate need to vote for things that have longer term ramifications. I'm sure you'll get plenty of conflicting ideas. But there'll be some ideas that may see no immediate value to the community, but have a larger pay off. To me, the obvious candidate there is Gen migrating to a unified encyclopaedia model (so no separate HE/CSE tables) which is underpinned by Unicode. It'll greatly simplify CA's technical debt (no secondary reimplementations, drastically simplified code page scenarios), and should allow you to turn around features significantly quicker because of it. But in the here and now for any customer? There's probably no real value for them, and I doubt it'd get many votes.
The challenge, I suspect, will be to weave those ideas into a larger, longer term vision. It's easy for me to armchair quarterback, I don't envy your staffing and resources issues.