The problem is that there is so many factor to take in account, frame material, thickness, wheels, quick release skewer, big chainring, tyres pressure etc. to name but a few.
Testing in labs does not always translate to the real world.
That was kind of the point I was trying to make, real world performance can be very different from just calculating performance of fictional tubes/spindles. But I really don't want to be embarrassed any more, so I take everything back that I said.
You're right, my bad. I shouldn't have used the words steel and aluminium. You are better than me.