If you succeed, you should probably be working at NASA. ;)
They were launching craft 50 years ago and doing the math. The computing power I have in my toaster is probably 1000 times what NASA had then. It'd still be a bit of a project to read up the academic papers and figure out how to approach the problem. But even a completely naive implementation would probably run more than fast enough. Especially if it's an autopilot feature for a proper simulation type game, there's nothing wrong with a "plotting autopilot course" wait dialog of a second or two.
It was a bit of a joke. I honestly believed him when he said that he already had a working solution, and I honestly didn't bother reading his web site beyond the first two paragraphs, and perhaps that was my problem. My first reply assumed he was doing something simple, like a Hohmann transfer. After alvaro posted a more realistic answer, I realized otherwise. Still, I didn't bother to read the guy's web site, and perhaps that was my problem yet again. Anyway, please consider that it took me more than 2 minutes to write my first reply, and even longer to edit it to reflect the new knowledge.
The key word that I was going on was optimal, and the link to the book that I posted seems to be for fairly new theory (~20 years ago, for JAXA and NASA missions). Perhaps I made a mistake when I assumed -- using my new knowledge -- that what he meant by optimal was optimal fuel usage in a chaotic environment, perhaps not. In any case, plotting a course with optimal fuel usage in a chaotic environment requires a bit more than just naivety, as the 220 page count of that book goes to show. Outside of the bounds of chaos (ie. a couple of two-body problems stitched together, such as the Hohmann transfer) where things are reasonably predictable in the long term, yes, calculating ahead to see where the target will be at later is not exactly rocket surgery. Outside of the bounds of optimality and determinism/chaos, where fuel is infinite and/or enemies are unpredictable even in the short term, yes, AI for chasing and evading is most definitely nothing remotely close to rocket surgery. ;)
Thanks for the input either way. I really did learn a ton of stuff from you, and I rated you up for it (yeah, thanks for those Intel fluid simulation links too). ;)