Another consideration: if White has a pawn on e5 and the possibility of using the g5 square, Black might well want to take on g6 with the f pawn. The reason is that a subsequent ...h6 is less damaging to the Black K position than ...f6, which would allow exf6. Here is an example given by Soltis in "100 Chess Master Trade Secrets":
The details didn't show up in the post: the game is Spassky-Petrosian from their 1969 match.
Opening the f file would be good if it were coordinating with something.
It's like someone calling a venue, and paying a fee to have their band play a concert... but they don't have a band yet. It makes no sense. First get a band together, then start looking for venues.
Sorry for the analogy, I'm a little out of it at the moment.
I any case, higher number of pawn islands can make both your mid game and endgame weaker. It's a good sign you're willing to think outside of the box, but until you're a little more familiar with the underlying logic, it's better to stick to principals.
Also, don't feel sorry for the analogy; I actually found it enlightening. This analogy resonates for me because I have been attempting to think "more outside the box" lately. Personally, I love the endgame and so I used to be constantly thinking "endgame." What I've recently come to realize is that sometimes you are giving up a dynamic edge to always play for such endgames. The element about the venues sparked my insight because I think this is the kind of thing I was looking for through this forum. If I can coordinate play then opening the file might be stronger, but if I can't do enough with it, then I am hurting my endgame chances.
I like how this is described, it sounds like a trade-off of attacking opportunities for now, versus pawn structure for a later endgame; it is sort of like an investment, but ultimately the chess player is trying to decide if it is a good deal or not (based on the position and if their play has compensation enough for the weaknesses)