Scientists, ballisticians, and countrymen, lend us your minds!
Guys, please read these thoughts & poke holes in them. My Dad was a chemist, and my nephew is one, but the scientific gene skipped a generation in yours truly
Hypothesis: a. Brits chose .380/200 for its deep penetration and blunt-nosed smashing effect when it hit bone, plus it "pushed right through" when it hit target. (NOTE: at this time, I don't know the barrel twist rate of Enfield/Webley revolvers, so cannot account for this factor yet.)
b. American police, etc., using the Colt New Police (.38S&W/200, with flat nose), probably had the same results. Ditto for those using the Winchester factory cartidge, with its very blunt nose.
c. American police, etc., using the "pointier" Remington .38S&W/200, probably were more likely to experience tumbling, as the pointier, even longer, perhaps less stable bullet tumbled when it hit the target, as it was only marginally stabilized.
d. Tumbling may also have partially resulted from use of S&W revolvers, with their slow twist of 1:18 3/4". Possibly the faster-twist Colts stabilized the bullets more fully, reducing likelihood of tumbling with any .38S&W/200 ammo. Likewise, faster-velocity 200g loads, such as the .38 SPL "Highway Patrol" load of c. 730-770 fps, may have been more stable & therefore less likely to tumble.
e. Possibly, changing barrel lengths from longer to shorter (i.e. to snubbies) may reduce velocity and stability of one or more loads to the marginal stability point, and cause a load to tumble in a snubbie that doesn't normally tumble in a 4" or longer barrel. Additionally, it is imaginable that a gun/load/bullet combination that is stable at close range, could become unstable as its velocity decays to a critical point over longer ranges.
f. Based on (a) thru (e) above, I hypothesize that the famous "tumbling" effect was more likely with "pointy" bullets at low vels (c. 600 or lower) when fired from S&W snub-nosed revolvers, whether in caliber .38S&W or .38SPL, and at relatively longer ranges. Conversely, that blunt bullets were unlikely to tumble in general, and even less likely at close range, when fired from Colts, 4" S&W's, or in higher-vel loads such as the "Highway Patrol" loading.
PROPOSED EXPERIMENTAL METHOD: attempt to obtain an adequate supply of both "pointier" bullets and blunt designs. Develop loads of c. 600, 700, and 750 fps with both bullet styles, to approximate .38S&W, .38 Colt New Police or .380/200 British, and .38 SPL "Highway Patrol" loads. Test each load from Colt, S&W and even Enfield/Webley revolvers against water jugs and wet newspaper, to determine which factor, or combination of factors, is more likely to give a tumbling effect. Test at close (7 yds), medium (25 yds) and long (50 yds.) range.
If we can succeed in identifying any real trends, we may have some decent answers even before we get our hands on British 1920s testing results and analysis. Accordingly, a modern-day shooter can optimize loads for his gun, because he can more reliably anticipate the effects associated with any of the popular loads/vels, bullet profiles, and revolver/barrel choices. For instance, one might learn to use a pointy bullet at low vel with an S&W snubbie if he wants the "tumbling" effect, whereas someone with a Colt may determine that "tumbling" is unlikely anyway, so he should load blunt bullets at the highest attainable velocities. If someone has only blunt bullets, he may forget tumbling completely.
If we can confirm my hypotheses, then it's time for more testing. Who out there has a Colt revolver in this caliber & would like to do some testing? How about Enfield or Webley? And a S&W .38SPL snubbie? Maybe some Remington (or other?) "pointy" 200g bullets?