I read through it and I actually have an issue with your point about bluffing frequency. I am not saying that it is incorrect--it is absolutely right, however I cannot see a practical application for that theory. There is simply no chance in hell that you are every going to have the type information you need to make these kinds of decisions in the way you describe.
Let's look at your example. You say:
"Let's say you are on the river and your opponent has checked. He's a decent player, but by now his hand is somewhat defined, while yours isn't. Looking at hand ranges, you are going to have his likely holding beat with 30% of your hands."
One of your solutions is:
"value bet half the pot with your 30% winning hands and also bluff with and additional 10% of your hands? From your opponent`s point of view the EV of his call is exactly zero."
This situation, while simple at first glance, is so convoluted and hypothetical that you can probably play for years and never run into even two which you can put into an analysis sample. Yes the math works out perfectly, but that's as far as it goes. First of all you are assuming a generic "decent" player without any personality traits, specific playing style, past history with you, or awareness of what has happened in the last few hands, who incidentally plays near perfect. Second you say, that you are going to have his likely holdings beat with 30% of your hands. This implies that this perfect situation you are describing will reoccur often enough to build a sizable sample, whereupon, you will be able (using your precognition I am assuming) pick out 1 of 7 of the hands that are behind his range to bet the same exact amount so that the math works out. Moreover, this situation is so perfect that which 1 in 7 of your losing hands you pick to bluff with does not matter.
This whole description reminds me of the econ courses I took, where the theories were all perfectly-structured, but so many of them were completely useless when you attempted to apply them to reality.
The whole process simply does not work the way you describe. Your decision whether or not to bluff is never based in such general terms. When have you ever found yourself actually thinking in the terms you describe? "Ok, so I am guessing here that I am ahead of his range and this situation is such that I will be ahead of his range 30% of the time. Ok, so to make the most of this I am going to bet 1/2 pot (or however much you decided was best, which amount you will base solely on the poker math behind the scenes) now and when this same situation comes up in the future, except that I am beat, I am going to randomly pick one in seven of them to bluff." Let's get serious.
Ok, you say, I am missing the point, this is used strictly for analysis of your game based on the history of your hands that you have already played. Fine, in that case, I challenge you find even TWO hands that you can put into the same category for this exercise, one hand where it's everything you said and you have the best hand, while the other hand is that same exact hand except that it's the inverse and you are behind the guy's range the remainder of the percentage of the time.
I am missing the point again, of course you can't be THAT precise with this. It's about the theory, so you have to make some allowances. But no, you can't do that. The math works out IF AND ONLY IF the numbers are precise and the situations are exactly identical. In the real world that does not happen. At the poker table in any given hand, you just have those 2 cards and they are either definitively ahead or definitely behind your opponent's range. If you think you are ahead, you try to extract value. If you don't you try to see a cheap showdown or fold. Your chance of a successful bluff is based upon a whole multitude of factors, including past long and recent history, the tendencies of the player, etc. etc. Even trying to group that hand into some category is an exercise in futility.
All that said, yes you can find out whether your bluffing too much, too little, or about right. But you do it by going back and analyzing your play hand by hand and looking where you went right or wrong, not by trying to employ some convoluted math exercise which completely disregards idiosyncrasies of each hand.
Please do not take this as a flame, as your article does give a lot of food for thought. I just strongly disagree with attempts to generalize real life situations into a synthetic mathematical model, right though that model may be in its own realm.