[hinkley]

I think part of that is because we are expected to put up with frustrating software at work because we are getting paid for it so if the work is more work than it needs to be, well just shut up and do your job.

Games are supposed to be fun. The money goes the other way, so the annoyances are taken a bit more seriously. The creator has to take that frustration on themselves instead of externalizing it, which is more work for them.

Add to that the delight aspect, and games try crazy stuff that would never fly with business software. 'Safe' doesn't get you many accolades so you're essentially obliged to try at least a few things. And the shelf life is lower so even if you fuck up you usually only have to hear about it for three years and then you can try something new. Once in a while, every five years or so, one of those UI gambles ends up being adopted by the industry, so while not the most effective petri dish for general UX research, it still gets some results.