Assume I own a game studio and someone (say, Louis Vuitton) wants to convince me to add a purse NFT to one of my games.
Unless they pay me to do so, I'm not going to dedicate the several-person workload (between artists, shader techs, QA, and production) necessary to implement that asset into the game for a single user unless they pay me to do so. This seems like a reasonable requirement on my end (and although there may be a company or two that might warrant the "free" work, the vast majority will not).
Therefore, Louis Vuitton now has a "cost" for their otherwise-free value-add NFT -- which can now be multiplied by the number of game studios they need to convince to add their NFT(s) to their game(s). Obviously, this can't scale to "any" game in the current industry (e.g. literally thousands of incompatible engines in use during any year's new releases) and also increases in cost over time as more games come out and/or older games would get updated.
This is also also for Louis Vutton -- I'm even less incentivized to spend more and more manhours implementing NFTs for smaller and smaller brands, and every brand that requests their NFT(s) added to my game(s) is competing in a zero-sum game against eng resources being put towards, well, real work that generates ongoing revenue for the company.
In the current industry, there would be literally hundreds of thousands of manhours required (if not more) to add a single NFT to "any" game, and still a significant amount of that to add a single NFT to "some" games. The primary beneficiary of adding NFTs to a game is Louis Vutton (with a value-add on purse sales) -- not the studio that actually has to do the hard work to implement it. So... why would they?