I'm not sure if the observation is right in the first place. If the question is about software for Maemo and getting developers to write for Maemo, I think it's a pretty impressive success story.

Don't forget that Maemo hasn't been available on mass market devices so far. Also, there's no "app store" where you could make money from applications. So what you could realistically hope for is that those who own a Maemo device may eventually develop/port what they need for themselves and make it available for the public.

We have ~500 applications now. That's plenty. Really. If I go to the Ovi Store with my S60v3 device, I get ~1000 Applications (including games). Now compare the number of S30v3 devices sold to that of Maemo devices... And add to that the fact that people who publish @ovi.com can make money from their applications...

My impression is we have a much broader range of software available than one should expect in this niche market. Don't forget that when application stores for other platforms list 10.000 apps, you can safely assume there's always 500 of them that do the same thing with a different UI. We may have only one or two apps for a specific task, but that's better than having 500 applications that make your N810 fart in stereo.