Android has a major problem: it's slow and project butter seems to be a giant patch rather than a decent fix. Ubuntu Phone OS follows the same open philosophy of Android, so it could eat a big share of Android's market if it's fast enough. Ubuntu Phone OS is based on Qt and QML which are really fast, although QML is an interpreted language. But here is the thing, it won't appear until 2014. By that time Google probably has rewritten the entire Android graphics stack to make it fast. My foresight is, therefore, that Ubuntu Phone OS will fail, exactly the same as others libre operating systems like Jolla, Meego, Mozilla's OS, etc.
I really like Ubuntu Phone OS: it has a desktop counterpart (I hope this means perfect integration) and it's based on a C++11 framework, Qt, that is complete, modern and open. Qt is comparable to first-class frameworks like Cocoa; with the addition of QML development is made easier and productivity is increased. I know that Qt is not the only thing that affects speed: it's placed at the top of a complex stack of libraries. I'm assuming here that Canonical will end up with a graphics stack that performs better than its Android counterpart, which probably is not very difficult. The fact that it will not ship until 2014 is definitely going to make the project fail, I think. Although if Google doesn't fix Android's graphics stack it will be a serious contender in the fight for getting more mobile market share.
To sum up, another opportunity missed...
