I don't think you're truly comparing apples and oranges. If you can point out a company about the same size as Microsoft, Apple or Nokia who is also publishing their work unencumbered by patents then I think we can fairly say that publishing work without patents takes longer. But your example misses the obvious detail that the companies working on patent-free technologies (e.g. Xiph) are small, relatively low-budget groups - even the Dirac project has been slow simply because it's a very small part of the BBC and has about three developers over its lifetime. This is probably related to the fact that these groups don't charge for licenses for using their products.
In reality, swift development is caused by large, well-funded teams, not by patent licenses. (And it's still not a guarantee of smooth sailing - look at all the companies now claiming to own patents on parts of the MP3 decoding or playback process...) Patent licenses are just coincident with well-funded teams.
Another counterexample: the CELT codec developed by Xiph.org is an excellent low-latency codec that competes with the best high-latency codecs in quality, and it was developed over two years by a very small team. Its core mathematics is unpatentable as the main paper covering the idea was written in 1975...
I agree with you on the point that patents improve the landscape for creators at the expense of technology adopters. Other organisations and governments are realising this too and I think we are starting to see a shift in the landscape away from encouraging pure creation to encouraging adoption.
(We need to keep the balance, however, because it's easy to argue that adoption is best served by a monopoly on who can create new software, and that's just what Apple and Microsoft would be happy to support.)