I'm new here.
Pretty much all I have to say about playing White against the Grunfeld is "see game 1 of the 1950 final candidates match, because Bronstein had the right idea, I think".
Whether the question is as broad as "Is the Feldercarp Attack good?" or as specific as "In the Hullgully Variation of the Potzgrebie Defense, does GM Hornswoggle's innovation at move 23 make it a good choice for someone rated 250 on lichess?", will I be frowned upon for rubberstamping each with "You can't think about these opening decisions until you're well-learned in the middlegames that arise from this option or that. And you can't play a middlegame worth a fiddle unless you can tell the difference between a favorable or unfavorable ending. If there's an ending on the horizon, how will you know whether to head for it by simplifying the position or complicate things by keeping the pieces on the board unless you can tell if the coming ending is good for bad for you. In other words, sell your openings literature for 50 cents per book (because the quarters are useful for parking meters and washing machines), and learn something useful, like endgames and tactics."
Whenever there's a question about an opening, I want to answer it like that. That's what I tell people in classes, lectures, privately, and so on. I have an equally long spiel about how handtrucking one's entire openings library to the board, and playing with open books won't help one bit. Because so what if you reach the end of your book line with that magical plus-over-equals? There's nothing in plus-over-equal to show if the resultant position is suitable for your style or temperament, and unless you're well-grounded in tactics, you're likely to hang a piece on the next turn, anyway.
I think opening discussions are poison for club-level players, and the many who think they're better than club-level. And since we're trying above all to be helpful here, I think it's most helpful to tell folks that they're barking up the wrong trees by asking about openings. But I reckon the community at large would not like it if, like I said, rubberstamped such responses.
I'm trying to earn good karma, because I have really dumb programming questions to ask in other places, and I genuinely want to help people play better chess (and that starts with as few pieces on the board as possible; first you learn with two pieces on the board, then you learn the positions with three, and four, and in about 50 years, you'll work your way up to 32 pieces, because those are the really hard ones... damn it, there I go again).