Hiya.  I think you're combining a game theory concept with an exploitative strategy concept and that's creating some confusion.

If we're betting bigger, our opponent is supposed to fold more.  Essentially, his minimum defense frequency (MDF, check out some of my first videos for a full explanation) is going to be lower as our bet-size increases.  But that doesn't mean he should now fold out all his bluff-catchers like you seem to indicate.  If he folds more than he should due to our larger bet-size, that's great.. our bluffs will profit a ton.  Sure, our value hands no longer make as much money, but our bluffs make so much that our range in it's entirety still benefits more as a whole thanks to him over-folding. 

If you want to start a mixed strategy of betting larger with bluffs (to get him to fold), and smaller with value (to get him to call), then you're applying an exploitative concept in hopes that your opponent doesn't realize what you're doing and you get to maximally exploit his mistakes.  To be frank, this strategy tends not to do so well in mid+ stakes online anymore, so I would refrain from making it your default approach. 

The reason betting bigger on the turn as a GTO strategy is likely better than betting smaller is because of the dynamics of the situation, given that we're quite polarized and playing against a very non-robust non-polarized range that models well as simple bluff-catchers.  Thus, we take advantage by maximizing the amount of hands we can bluff with (with a larger bet-size) and subsequently, minimizing the amount of hands we have to give up with.  In theory, he should call enough to make our bluffs indifferent and this strategy will maximize our value hands (because our bets were bigger and we will make more volume overall).  However, in practice we'll often encounter players that will have trouble responding well to the strategy and will overfold or underfold, both resulting in us making even more money.  In summary, if we're balanced, we don't care if our larger bet-size will "likely result in a fold from our opponent" or if our opponent will be curious and look us up too much, either mistake will be costly and ultimately there's nothing he can do but try to respond GTO to minimize our value extracted from the situation. 

Sorry!  Too busy eating 6k calories a day and playing HSPLO to make good NLH theory videos anymore.. :(  But the RIO pro line-up is becoming increasingly awesome so I would definitely recommend checking out videos from a lot of the other coaches.  

Good point.. I think my line of thought at the time was that K-high was strong enough to check back because it could win at SD some significant % of the time and that we'd rather bluff worse hands that will never get to SD and win.  The reason we bet the KQ/KJ hands is because they work nicely as robust hands that can work well betting over multiple streets as they (1) sometimes make the nuts, (2) sometime make an okay hand that likes to check-back and get to SD, and (3) sometimes will work out as a great bluff candidate.  I think all of these things are debatable and its these subtleties for range construction that are important to be thinking about but it's difficult to come to concrete solutions.  My NLH experience over the last few years is lacking so it's possible there's a better way to distribute these hands.  But the key lesson to take away from this is that these little tweaks on hand choices for bluffs is not going to be as important in the grand scheme of things as your overall frequencies and ratios are.  The goal is to be balanced with appropriate ranges to match our bet-sizes and fit the situation well, and from there we can debate which hands are going to be best chosen to fill-up our bucket of bluffing hands.