Not completely, but it is wrong enough to forget and relearn from scratch.
The river is easiest to analyse. Then, to a very good approximation, the odds you offer your opponent dictates your "correct" value/bluff ratio. If you offer him N:1 to call, the odds against you bluffing should be N:1. This means, for a 50% bet with N = 3, that you have 3 value hands for each bluff. That is, you bluff 25%. With a pot sized bet, it works out to 33% bluffs. Whatever villain is/should be doing is of no concern. Generally, smaller bets = fewer bluffs. The logic is easy in this case. With small bets, it is cheap for villain to call, he get good odds. He will, and should, be calling a lot. With N = 3, he needs to nip you off 25% of the time to be BE.
Adopting these calculations for earlier streets is called the "one-street alpha approximation". The numbers you quote for F/T/R are probably from a bluff stacking model. They are i m o quite wrong since you'd be c-betting about 150%, turn barreling close to 100%. Bluff stacking models have never been confirmed mathematically (and believe me, this is a horribly difficult problem, it requires a mathematician's professional effort, and I doubt that it is within human reach to prove the numbers).