I think it's important to remember that LiMux, the Linux project in Munich, didn't really fail because the software didn't work.
The city had commissioned a study that blamed bad implementation, bad internal cooperation, and bad administration. It explicitly did not recommend that the project be shut down.
Unfortunately, this recommendation was ignored by the mayor, who was previously responsible for convincing Microsoft to move its German headquarters to Munich and who calls himself a "Microsoft fan".
So it's probably worth noting that the success of such large projects doesn't only depend on the viability of the software. It's also very much dependent on the lobbyists the project is up against, especially in the public sector.