I don't see what makes video codecs so much different/more special than other areas of software development. Surely you're not suggesting that software development was stymied and slow before patents? Surely you're not suggesting that the only motivating factor for interesting or innovative (or rapid/efficient) software development is patents? That is demonstrably false. So why are video codecs unique in this way?
For what it's worth I've worked on a LOT of software projects for companies both big and small, and never once in all those years was ANY decision about what to build or whether or not to build something EVER based on whether we could get a patent or not. It's just not true that that this is something companies think at all about, or depend on in revenue forecasts etc., before they commit resources. Similarly, I've NEVER been involved with or even heard about a situation where development was stopped because it was discovered that the work was not patentable after all (obviously stopping or changing work because it was discovered that the work was or might have been already patented is quite another thing).
What really happens is that mid-to-late in the software development cycle managers ask the more senior developers to consider whether anything they've been working on might be patentable, and if so to help the legal department fill out a patent application. Most of the time you don't hear anything about the patent until you've already moved on to completely new things.