This doesn't feel so much like graduating from the concept of indifference as it does repackaging the incentives of exploitative play. Bluffing combinations that are meant to make V indifferent to calling with bluff-catchers is 0 EV under the assumption that V is assuming we are playing balanced and so is playing to be unexploitable; but if V is calling too frequently, then we understand that these 0 EV combos aren't 0 EV anymore. It's clearly -EV to ever bluff into someone calling with 100% of bluff-catchers. But we knew that already; that isn't using external factors to nudge a 0 EV play one way or the other, but is just exploiting the tendencies of a player whom we can predict.

Or, take as an example something you gave in your video; making a slightly -EV play to demonstrate that you aren't a nit, and to show you have big bluffs in order to increase the EV of value hands later. But if we have the image of a nit, we're already, in that first hand, expecting more folds than equilibrium, all else being equal. There's no reason to deliberately deviate from a decision we, in the moment, perceive as => 0 EV just for the sake of making a new and different table image. Whatever our table image is then and there, there is presumably some way to extract EV out of it given how people will heuristically react to it.

Although I get the general point of the video, which is that you can take how players view you into account to inform your exploitative plays, the way you frame it is incorrect from a theoretical standpoint, since:
- When making an exploitative play, you don't really need to do it only with the hands that were indifferent in equilibrium, say, against really bad players you could open 83o on the SB profitably while it wouldn't be anywhere close to indifferent in equilibrium.
- Your hands EV doesn't exist in a vaccuum, it depends on your whole range, therefor, if you could somehow input into a solver all of Villains tendencies, reactions to your image, etc... the maximally exploitative strategy that it would output would still have indifferent hands that it would mix, they would just be different hands than in equilibrium.

Interesting video.

If a -EV “mistake” is the highest EV move in the long term due to metagame tactics, then it’s not actually a mistake but the correct exploitative strategy and, therefore, +EV.

Reframing the timeline from one hand to a lifetime of hands resolves the paradox of -EV metagame moves.