221540
Unless you're coding it yourself you can't really make assumptions on how long it'll take to do something. If they wanted to do it right it would take a lot of work. Like how some bosses are really easy to beat but tedious because they gave the boss 1000000000 HP to try to make it hard? They can't just raise the levels. If they want to make a real hard mode they need improve the AI of the trainers which is far more complex.
I don't really need to code it, myself; I have a pretty understanding of how Pokemon works. Really, they wouldn't even need to change AI, they could easily steer the AI with movesets. Pokemon uses items and moves as they deem necessary, and the RNG for what move will be used is chosen based on the move that is chosen by the player. -Not optimally, mind you. There's a good bit of randomness in there both to avoid a future-seeing cheating A.I. and to make it so that the AI isn't too smart.
So I can see where the trouble would be. The A.I. isn't dumb, exactly. It knows what Pokemon is weak against what but it doesn't learn from the moves used (which, as generations go on, is more and more reasonable). But tightening that would require something of a learning AI, something that changes its tune to right its wrongs, and that sort of algorithmic concept isn't really there in Pokemon. They have little to work off of in that sense.
And I see what you mean, I haven't seen the code so I can't say what its like, but I think estimating and assuming are two very different things in this case, and I think, on a conceptual level and even a logical level, assumption is a perfectly fine course of action, especially considering GF's current size.