That's really not a very good idea. Fly is a warp system, first and foremost, a necessity in any game as geographically large as most core Pokémon titles. There is no good reason to needlessly overcomplicate something that's there largely to make travel easier and faster. Can you imagine just trying to fly to a city to use the Pokémon Center there and having to deal with some crap like a minigame or more trainers before getting there?
More choices = more confusing = harder for the target demographic to play.
I realize some people think it's fun to theorize and brainstorm endlessly about what they think should go into a Pokémon title so it can fit their specifications exactly and I'll put $5 on at least one person shouting "STOP CRITICIZING OUR IDEAS!", but consider two things just the same:
1. If you're posting on this board, then chances are excellent that you are beyond Pokémon's target demographic, which is and always has been boys ages 6-11. If the games don't seem as fun or as hard as they did at one point, it's because you aged and your tastes matured (in theory), not because the games changed.
2. We've had four generations of at least three core titles a piece, and even as the handheld technology has advanced and improved to allow for new gameplay mechanics, there are clear and defined sets of elements, goals and mechanics that have remained unfailingly constant over the years. To call for change to those sets now is putting energy towards something that won't happen.
STOP CRITICIZING OUR IDEAS! - Spend it wisely.
I do mostly agree with you though. Pokémon's target demographic is younger then a lot of the members here are. I think that's why a lot of people consider the G/S/C era to be the best one... It's when they were at the ripe old age that Gamefreak had in it's sights.
I would like to know where you got the "6-11 year old boys" statistic from though, as I can only assume it's an assumption (albeit and agreeable and educated one)?
To be slightly more on topic, avatar customisation would be quite nice, but there's a lot of sprites they'd have to program, and frankly, a lot of the Pokémon sprites have enough errors already, so I'd like to see them fixed first.