Your analysis is excellent. I see people using every upward deflection of the HL curve as proof that the URR is larger. This is wishful thinking. The system is closed as far as the oil is concerned so there cannot be only positive deflections. You demonstrate that in all likelihood the URR is estimated best by HL data that arises during the normal production regime and not forced transients when production stalling starts to become apparent. These forced production regimes show increased production and delay the peak but they do not create new oil. What you extract faster today, you will have less of tomorrow.

upward deflection of the HL

Of course if we had enough HLs for enough fields "deflections" could be studied -- average length, degree of deflection, etc.

It would be nice if some HL curves were shown for truly end-of-output conditions. In the Texas case I expect the HL curve to start heading down in the next few years for a URR below 70 Gbl instead of keeping at its 2006 trend towards a URR higher than 70 Gbl.