You must have missed the thread the other day about open source video games that used to be commercial. Psychology departments can run experiments on people where they see what the effect of using Word Perfect is, or otherwise use it as a learning task.

So, for example, you can try to design the fastest learning protocol for a given set of functionality being acquired by 90% of subjects. It's fairly safe to assume that most college students don't know Word Perfect, so you should be able to get good results doing A/B testing.

You could do experiments like: do people learn Word Perfect faster if they take a 30 minute nap beforehand? Do they learn worse if they do 30 minutes of intense exercise beforehand?