I too am mystified by Gnome and Canonical. If they want to compete in the tablet/mobile space then they need to be good at something. Anything will do such as price, freedom, apps, developer productivity, distribution, popularity, localization, tools, privacy, security etc but I can't name a single thing. And the competition is constantly improving - being better than Android/iOS today is useless if you ship in 3 years. Not to mention Mozilla/Firefox are also messing in this space diluting effort and attention even more.
The only thing Gnome and Canonical have to show so far is a homescreen/launcher which is by far the least useful/interesting part of the tablet/mobile experience.
On the being constructive side, what they should be doing is what happened to the GNU project. Do not try to take over the whole world with your stuff all at once. Instead run parts on existing platforms, and provide alternatives or replacements for the other components until it is possible to run completely "native". This way the development effort can be used most effectively, feedback is quick, and you have a working system from day one. What Gnome and Canonical seem to be doing is following the model of Hurd!
> The only thing Gnome and Canonical have to show so far is a homescreen/launcher which is by far the least useful/interesting part of the tablet/mobile experience.
I wish the general public agreed, but the number of awful websites for new desktop wallpapers or icons or mouse pointers or other trivial customisations tells me that a lot of people think the 'home screen interface' is super important, and they don't care (or even know) about other more important things.
Canonical and Gnome have destroyed my previous arguments about the benefits of benign dictators or of directed committees. There's some gentle backlash against Ubuntu (a little bit is from people who know what they're talking about, a lot of it is from people who just don't like the popular Ubuntu) and there's similar backlash against some of the Gnome stuff.
I really wish that tablet / phone / touch interface things were spun off into separate projects, to allow people to continue polishing existing projects rather than focusing on monumental change. That'd allow people to continue using one interface on desktops and the other on tablets. Because they are different, and have different needs, and you can't really kludge one metaphor onto both devices.
That's what KDE has done: the big rewrite of KDE4 let them use the same libraries to power two different shells, for distinct purposes. That's a good way of doing it. Though it does not solve the problem for the rest of the applications.
Once users are happy with everything else, then they turn to customising home screens. If the entire system was only a home screen no one would touch it! (And arguably many are doing personalization rather than customization.)