The problem with people who think of ideas and don't try to execute them is that they can't incorporate feedback. At best you get a second-hand explanation of why the magic thing can't be built.

You also end up with an attribution problem: it's easy to say what the big idea is, but it takes a lot of work to make it happen. Why give credit to the guy with the big picture, when the big picture is probably something the guy who implements it had as well? Devil is in the details.

You also don't get good ideas if you don't try to implement them. The implementation process is a creative process, which btw your ideas guy might not recognize, leading to issues.

I've had ideas guys where I worked before, and it was toxic. My view is basically that ideas guys are a kind of leech: they take credit for positive things that are obvious, they come up with excuses for things that are negative (team failed me boohoo), and they don't generate better ideas than people who are doing the implementation.

My experience is in quant trading, where you have people who can code, and sometimes people who just fluff about strategies. With the randomness and noise it's easy to be fooled by ideas guys, and it's taken me a long time to just dismiss the concept entirely. I actually just got off an interview where we discussed this problem in the space, and we agreed it's absolutely essential that people can code, otherwise their trading ideas are worthless.