Also, correct me if I'm wrong but, it seems to me that bitcoin doesn't need much more developpement, does it?
I feel the opposite, the more popular bitcoin becomes, the more work needs to be done.
Security, robustness should be high on the priority list. Unlike other FOSS project where financial impacts are indirect, bitcoin software leaves very little margin for errors. A successful hack can cause massive and instantaneous damages.
It will also be nice to see implementations in other languages such as Java, Perl, Ruby etc. Having a (much) larger developer base to contribute or review the code will help.
Some peripheral projects mentioned on this forum:
Bitcoin point-of-sale (POS & SDK)
Bitcoin account hub (AH)
Mobile clients
Trading API
Web of trust integration
...