"How useful is it that such schemes must necessarily be very public and that might ensure that two teams of developers don't work on similar projects that end up with too few purchasers?"
What's with the pessimism? Come on! Give it a chance - there are serious software holes now and everybody needs one or another addressed. What good is it if people have to use another, less liked, OS to get a job done?
What hurt can it give if users simply vote for an app they'd like released, like the article suggests? It's then up to an ambitious programmer to see if it's realistically manageable. If a given program does not convince, users can be free to not buy it, but since there's quite a decent software standard in RISC OS I have plenty of faith it'll work out. I'd say if a lot of users vote for a particular software project, it should provide particular motivation for a programmer or even team of programmers.
I'd vote for a port of GAIM, ( [link] ), since R-Comp has again reverted to silence regarding their (apparently) stagnant Grapevine project, despite hopeful expectations a release could be made around Christmas...