Peter,
Thanks for some numbers.
(For anyone following this, note that in 1953 everyone still believed copper crushers were reporting pressure accurately, so where Peter quotes their psi, today they would be called CUP.)
SAMMI lists both a 4" revolver type test barrel (with 0.008" barrel/cylinder gap equivalent) and a 10" one-piece barrel (like a Contender barrel). A model 27's 8 3/8" barrel would have about the same length to the recoil shield as a 10" Contender barrel has to its breech, given the different barrel measuring methods used. So I would expect the velocities for the model 27 to be about what SAAMI lists as standard for their 10" tube.
SAAMI doesn't have a 153 grain bullet listed, but for either a 158 grain soft point or semi-jacketed hollow point, the velocity standard is:
1220±90 fps for the 4" revolver type barrel wit 0.008" barrel/cylinder gap.
1600±90 fps for the 10" one-piece barrel.
So I'm not convinced there's a difference in the standard that is real. I looked at a box of Winchester Super-X 158 grain JHP's and one of SP's, and it claims Muzzle Velocity to be 1235 fps, which is certainly to SAAMI's 4" test barrel specifications, so it doesn't look like it's loaded down any.
Your gun probably is not as tight as a 10" SAAMI test barrel and maybe not as tight as the gun HP White used. The SAAMI standard requires each round to be conditioned at 60-80 °F before firing. It has to go through the "SAAMI twist" to keep powder over the flash hole for firing. I don't know how much your testing deviated from that, but differences could certainly tell. I don't know how accurate your chronograph was that day. I've see two side-by-side disagreeing by 200 fps on one occasion.
The ballistic lab sounds like a good project. A couple of things I've noted over time: Denton Bramwell's stats indicated strain gauges are actually more repeatable than either conformal piezoelectric transducers and much more than copper crushers. Jim Ristowe at RSI, the maker of Pressure Trace says 60% of his sales are into the industry rather than to private owner, so it seems to be increasingly a less expensive alternative that is accepted. The main limit it has is that it is 8 bit rather than 10 bit resolution, like an Oehler model 83 commercial test instrument.
The CIP claims its channel piezoelectric transducer system is accurate within 2%. I don't know how they arrive at that, but they certainly are more consistent in the numbers they produces as compared to copper crushers than the conformal transducer is. You do have to drill holes in cases to use their system, though.
Bryan Litz has tested a lot of different chronographs and determined that 15 fps is about as close as the optical ones get, real world. He is currently recommending the long-sabre version of the Magnetospeed product. I keep waiting for the Labradar Doppler instrument to become available and prove itself, as it should be the most accurate, in theory.