I guess it depends on what you mean by rock solid. If RUO 2.0 is rock solid then packaging it up and releasing it would probably be a good first step before taking others. If you meant it in some other fashion like the core is solid, but maybe various other scripts need work still, or some other facets I as a newbie to RUO and the community don't understand then just ignore me of course.
But before moving off to other projects I definately would like to see RUO 2.0 released as completed, particularly if it is rock solid and ready for said release.
That said, I agree with many others that a non-OSI client is desperately needed. One that can take full advantage of the protocol and lift limits that OSI imposed on it's client such as items limitted at 0x4000 max, wearables that exceed 1000 and keep their animation while sitting, more than the current number of layers and more configurability in their drawing priorities. Configurability of sitting in chairs, configurability of dealing with containers. All these features are 'locked' by OSI because they only need to deal with their own server and don't need more, but most shards that have a lot of customizability need these things.
There is another client in development by the author of Mulpatcher, which took some steps towards this. His client breaks through the 0x4000 max barrier, and he spoke of making more layers available for his client, but I think his attention has been drawn to other projects or perhaps too consumed by Mulpatcher to continue.
PlayUO seems really nice, and only missing a few features like sitting in chairs, and the ability to break these barriers I spoke of, but it seemed functional last time I played with it. Unfortunately it also hasn't seen a new release in approaching two years (3/05). Projects that are for one reason or another not progressed by the author would probably do well to be made open source and allowed to fork off to various others. Eventually someone who knows what they are doing and has the enthusiasm and energy for the project might take it up and bring it to new heights and we can all benefit from that.
But a solid flexible client for RUO and all UO Emulator servers would be a next big leap that would make many of us happy and one step closer to abandoning OSI for anything but their art.
As for trying to develop an entirely new custom game, I think the big problem there for most has never been a lack of ability in the programming end, but a lack of ability in the artistic side of matters. Whether you are designing art for 3d or 2d, when you lay before someone the fact that they have to create thousands upon thousands of high quality items and animations and such to make the world equivalent in depth to UO, it usually is the issue that stops most such projects and attempts dead in their tracks. The graphics in UO may be outdated, but one must admit, they are decent quality in general, and finding teams of people who could produce the sheer volume and quality of such art for free(and sometimes even paid) would be daunting. Not impossible perhaps, but to date it has not been accomplished by any project I have seen. So unless you can gather the artists and solid commitments and work from them prior, I wouldn't even bother with working on a custom game that wasn't basically an emulator or at least leveraging the art of some other commercial game that people would have to buy to use it.
Of course, all the above is just my opinion, and not meant to ridicule anyone or their attempts to do anything or their decisions not to do something or other. What I would like done by you? Keep supporting the RUO platform, bring RUO 2.0 to release at some point, and perhaps looking harder at getting PlayUO going again.