I don’t want to pile on, so I’m trying to make this more constructive than just telling you you’re wrong. (not that I’m saying anyone else is doing that)
The videos you linked to don’t explain how to calculate a card’s position relative to the power curve, they just assume you can plot a given card’s power against its cost and compare that to the power of an average card of the same cost.
The “power” of a given move can be stated in terms of “utility” in AI and economics, so if you’re interested in this I recommend you research utility functions and game theory. The raw stat points of a card can be a useful heuristic, especially when you’re actually in a game and have a specific scenario to judge. But for what you’re proposing, I think the best solution is just play testing.
There are basically two types of utility: cardinal and ordinal. You’re method is cardinal, meaning it has a number of utility points. For example, Kindred Hunter has 12 utility points for 3 mana. Ordinal utility is just ranking your options from best to worst.
Ordinal utility is what we’re most concerned with, meaning we want to make the best move available. You’re trying to come up with a cardinal utility function in order to better judge cards. All I can really say to that is “good luck.” And I mean that sincerely. I hope this was helpful to you developing your method, or at least just interesting!
edit: I would also recommend this video by the same people about using abilities and special powers to make it harder to judge a card/character’s power.