Thanks, BCD. You have given me cause to think.
I was, in fact, just repeating the commmon wisdom which is, as you point out, not entirely correct but has enough truth that it has always worked.
Thinking about it, it makes plenty of sense that with a max of 15 psi (OK: 14.696) in the reverse direction that you can still seal appropriately regardless of the direction of the chevrons. I once had dealings with a valve that has spring-energized TFE Chevron packing and the knucklehead that designed it put the spring ON TOP instead of underneath. At about 150 psi the spring preload was overcome and the packing set displaced to the top of the stuffing box, and leaked. Spring UNDER would have been right, and would still have provided enough preload for vacuum.
If you DO get any packing leak on vacuum, it doesn't take much of a leak to make a difference.If you're trying to get 10 to (some negative exponent) torr, and there is any leak at all you won't maintain vacuum. Maybe it's the customary double packing set with lantern ring that protects from leakage. Certainly if the LR is supplied with a sealant (gland seal water in the power industry) it's more likely to seal and a leak of the sealant will not degrade system performance.
By your criteria, the performance of cup and cone packing is then equivalent. Both are virgin TFE or some loaded variation of TFE, Both are energized to seal by compression of the gland into the stuffing box( compression causes the id to be reduced and the OD to be expanded) , with supplemental process loading that assists sealing by providing additional compression. TFE chevron has reasonably delicate sealing edges that are a bit more subject to dirt and or poor-stem-finish abrasion, and TFE Chevron and C&C both must be assembled and installed as a packing SET in a particular sequence, not just random rings loaded into the stuffing box until it's full.