Ok Tom…so I understand what you are saying and agree there’s a bit of logic to it.
But you are implying that proper bike fit (and we are talking tri bikes I assume rather than road bikes) all hinges from the vertical axis rather than being relative to the body of the person on the bike. And then you are arranging the touchpoints from there.
No. All I’m describing is what the OP is asking about, i.e. when switching to shorter cranks, what adjustements to the other touchpoints need to be made to keep the relationships between the saddle, the bars, and the foot in the power stroke the same?
I’m getting the impression that you’re over-thinking this…
It seems that this would thereby disparage the concept of rotating the body forward around the bottom bracket which is how we always think of obtaining a good TT position…because darnit, the knee would now be more forward relative to the pedal! And we can’t have that!
Not at all. See the explanation above. My advice was given under the assumption that the OP has his touchpoints currently where he wants them to be and is merely looking to gain some “room” between his leg and his torso at the top of the pedal stroke. My advice for how to move the other touchpoints had nothing to do with the original fitting exercise.
If you are to be totally consistent in your advice to the OP, you will have to also have him move his elbow pads UP to complete your exercise of basically rotating him backwards around the bottom bracket.
I did say to raise the bars, didn’t I? I didn’t explicitly state to raise the pads, but I’m assuming they’re attached to the bars, which is pretty reasonable, no? 
I would assert that you are wrong to keep the leg to pedal relationship the same. First, on what axis would you be maintaining it? When interfacing with a circular crank the angle of attack versus vertical means nothing. It’s the angle of attack versus some point relative to the body that is meaningful. And with a shorter pedal it’s maintained by moving the seat foward and the elbow pads down.
Again, you appear to be overthinking things. The important thing about the pedal/foot interface is the path that the pedal takes through the power stroke (downstroke) portion of the pedal cycle. The “return path” on the upstroke isn’t as important. The power stroke is close enough to being a vertical that the adjustments I described above are perfectly acceptable using horizontal and vertical references. Being a practical sort of engineer, sometimes being close enough is…well…close enough 
The point of the adjustments when moving to a shorter crank (and not changing the touchpoint relationships) is not to rotate about the BB…the point is to translate the body relative to the BB in the longitudinal plane of the bike so that the relationships to the touchpoints (with the pedal touchpoint being at the point of max torque application) are kept constant while also keeping the orientation of the upper torso relative to the horizontal plane consistent.
The whole point is to not *just *make adjustments to saddle position when changing between crank lengths. That’s how you end up with people saying things like “I swapped from 175s to 170s and it just didn’t feel right”.
If one wants to make other adjustments (for example, lowering the angle of the torso relative to the ground) that moving to shorter cranks might help accomplish, then that’s an slightly different issue than what the OP was asking about.
Of course you may still be right. 
Of course
Then again, if you search enough, I think you’ll find cases where I’ve stuffed the direction of adjustments before…but, I’m pretty sure I learned from the pain of those experiences enough that I’m not making that mistake again here…I hope… 