There are few misunderstanding in this question. Let me explain.

Starting a Linux distribution is A LOT of work. So much work, that most attempts on this fail.
That's why most new distributions use an existing distribution as a base. Ubuntu and Maemo started basing on Debian. Mandriva and Moblin started basing on Fedora/RedHat.
MeeGo starts a bit differently - by merging best features from both Maemo and Moblin. Yes, what is visible at the first glance is RPM packaging, and a lot of packages taken directly from Moblin (as these were already there). But if you look at the architecture, it's basically a copy of Maemo architecture with bits of Qt replacing here and there.

Ergo: MeeGo is not starting from scratch. It merges best features from Maemo and Moblin.

Secondly - maintaining a distribution is a lot of work. More than you can imagine before you try it. Maybe a more than starting a distribution. That's why most distributions die after some time.
Yes, CSSU is a great effort to update the current Maemo5 state and it is successful in bringing a lot of new features, but updating the base system to current state is a mundane task which people get tired of. Especially if other distributions are doing it already, so if you base on them you get core updates "for free".

Thus Cordia decision to base on MeeGo Core. We have an updated core system, with toolkit and UI we like.

I guess it really depends on the viewpoint you're at.
End users are interested in new features and don't really care about the backends supporting the features, to the point of being oblivious that the lacks in backends cause the lack in new features.
Developers are interested in in APIs enabling them to do cool things with less effort and do not mind turning the whole system around with no apparent visible changes.