I read somewhere that if we 2bet and are facing a 3bet from a range with some 'bluffs' in it, we should continue in a way that makes his worst bluff 0EV. However this implies that we still allow all of his other bluffs that are just above his worst bluff to be +EV against us.

Why are we not trying to make some of his worst bluffs -EV for him so that in the future he reconsiders doing them against us?

Plus not only that but all his 'value' hands will be getting plenty of action from us too and printing even more money than his semi-bluff hands are, so where is our profit coming from with our continuing range if it only breaks even against our opponent's worst hand?

Again, you're mixing different concepts together.
When you're on the river and you call enough to make villains bluffs indifferent, your calling range is not indifferent versus villains bluffs, whenever he has a bluff and you call you win the entire pot. Villain's bluffs are indifferent against your calling+folding range.

Preflop it works way different because of a gazillion reasons like multiway, hands having heaps of equity... and probably more reasons that I can't think of. In practice you just have to make sure that you can't get 3bet by trashy hands, and focus more on what your EV is with each hand rather than what the EV of villains bluffs is

The idea that you want to make villain's worst bluffs 0EV is a toy game concept. It applies when villain is perfectly polarized to the absolute nuts and air with NO equity.

Preflop "bluffs" have equity, and preflop "value hands" aren't unbeatable. There are multiple players and multiple streets of betting. Some hands will over-realize equity, some will under-realize. So this toy game concept doesn't work well here.

Here's an example how how you would correct apply this:

Villain shoves river with the absolute nuts or complete air. You defend just enough to make their air indifferent to bluffing. If you defend less, they could bluff any two. If you defend more, they can stop bluffing and print with their value.

Because then they could stop bluffing and exploit you.

Their bluffs are break-even against our range (when considering fold equity). Our continue range is not breaking even against their bluffs.