To better apply this to the given example of OP, the exec's idea works on paper in the sense of creating a new inflow from whales who would, for some god forsaken reason, buy campaign mech skins. However, it completely fails to consider the backlash from fans and, ergo, resulting losses.
It's more a fact that they unfortunately do, but only that. Basically, even in the given example, they can crunch and run the numbers, but they don't consider the long-lasting impact from, on, and by people. My last job had the same issue in driving out the old guard while retaining the layabouts.
My old man described it more as a combination of nepotism and academics without experience. Basically, having worked in that transition between executives who built the businesses and the people inheriting it, none of them actually understood beyond theory, if any, on how to run it.
Like stuff with AI bgs? I guess it's just a persistence game then, because if so I get some of the same stuff but still avoid the slop.
Algos run on passive observation, but get knocked into overdrive by active rejection. Just stick to it, should work eventually.
Just hit "Do not recommend" and "Not interested" enough and don't engage with AI-related (negative or positive) content, and the algo will get the idea.
As someone who's been on both ends, it turns out doing the bare minimum of upholding your promises and keeping your customers updated is rare.
I've had a guy who berated me on stream for asking if he'd get to my sketch comm after a year. Needless to say, he lost a lot of customers that day, lol.