Knockdown Power, ICSA, Taylor KO Factor, and the plethora of other empirically derived numbers all have same short coming. If they work at all they only work for a very small range of values/conditions since they are not based on hard physics principals.
Correct. As you say, assuming they work at all.
Energy is useless no matter how you look at it.
Energy is certainly a useful quantity. Since it is a fundamental physical property of moving objects, it has real-world meaning and therefore value for those who understand what it means. The idea that it is simply a marketing tool is not only incorrect, it is a ludicrous claim.
https://en.wikipedia.org/wiki/Kinetic_energy
http://hyperphysics.phy-astr.gsu.edu/hbase/ke.html
https://www.livescience.com/46278-kinetic-energy.html
As early as 1722, it was proven that the amount of deformation caused by a ball hitting soft clay is proportional to impact velocity of the ball squared.
It is true that it can (and has been) overemphasized and that leads to confusion and misunderstanding. It is also true that many people don't understand it and what it means. However, underemphasizing it, or pretending it has no meaning or value will also lead to confusion.
I will agree that it is not possible to compare caliber performance across the board using nothing but energy values. In fact, I'll go a step farther and state that it is not possible, and I don't believe it ever will be possible to wrap caliber performance, across the board, into a single number that tells the whole story. It's more complicated than that.
Bullets aren't rated to expand at a certain energy level. They're rated for impact velocity.
Given that energy is dependent only on velocity and bullet mass, and given that bullet mass doesn't change from the time it is manufactured until the time it is shot, the bullet's rating can be stated either in terms of energy or velocity. It's just that stating it in terms of energy would require the consumer to perform a reverse calculation to determine the impact velocity. In other words, even though it is impact energy that is the primary factor in bullet expansion (bullet construction aside, and assuming a homogenous target medium), quoting the rating in velocity makes it much simpler for the consumer.
It is only viable for comparing big bores to each other.
To the extent that the TKO has any value, and without implying/agreeing that it does, I would agree that any value it does have is exclusively for comparing elephant guns.
That said, I think Taylor himself would disagree with your claim that it was "never meant to compare small bores to big bores".
His table of "Practical Striking-Energy" values goes down to calibers having bullet diameters as small as .240 and he comments that he included the "value of certain small bores to show that they cannot be considered safe weapons to take against dangerous game at close quarters in thick cover". Clearly he believed that his "Practical Striking-Energy" values provided a way to determine the relative performance of small bores to big bores.