My boss George is going to shoot me for tackling this; but its Monday so why not.
Thanks for the kind words. George doesn't offer HP/TQ ratings (We have customers in road racing, on the street, LSR, off-road, drag racing, and one twin-engine, blown Vincent motorcycle streamliner). Each of those applications has very different requirements for the gearbox. Even Richmonds ST10 had ratings way less than 400hp across several different gear ratios.
That being said; the stock Muncie M21 was designed around 1961 and debuted late in '63. It would have been designed for the engines of the day, and primarily for the Corvette (close ratios were designed solely for road racing at the time). '63-only Muncies are kind of weird, and definitely were designed before the 360+HP 327s were on the road. 64-65 Muncies have a skinny 7/8" countershaft pin and thin synchro rings. '66-70 Muncies use a 1" countershaft and sturdier synchro rings. '71-74 Muncies use the 1" countershaft, sturdier rings and upgrade to the 26-spline input and 32-spline output.
The problem with all Muncies is the case; the OE geartrain is far stronger than the factory case and weird, narrow 307 ball bearing.
I would say from a conservative point of view, every day driveability and 30,000mi lifespan...I'd be comfortable using a 66-70 Muncie with around 400hp (drag racing with slicks not included). I'd bump the fine-spline Muncie only slightly higher...because the 32-spline output was done to harmonize around the TH400 architecture (not lack of strength), you're still using the undershod casting with the skinny bearing. The only REAL bonus is the 26-spline input.
An OE NOS geartrain in our case is probably good into the 450s (Assuming M20/21 on the street). If it was an actual M22 gearset from GM, genuinely NOS, marginally higher. Its important to remember that a 50-hp difference between 500 - 550 is a lot different than 250-300 in this case. Also of importance is the fact that 2nd gear in a Muncie is stronger than 1st (remember it was designed for road racing).
At some point you run into the next weakest link; which will be the factory midplate.
Now; our stuff:
I routinely have customers with a complete M23 gearbox pushing between 600-750hp/ft-lbs on a case-by-case basis. We routinely spend a half hour or more reviewing each application and assessing the customers needs and wants. If a customer was willing to live with routine preventative maintenance and had a well sorted out clutch, suspension and tire package; with attention to bellhousing concentricity and parallelism...you could easily push it higher. If you put an M23-style 1st gear next to an M22-style 1st gear...the M23 design is probably 3x stronger. That borders on salesmanship; because you're still left with the underlying structure (case casting, ball bearing etc.) We are probably looking at 20-25% more strength as a system.
The M23 series uses a hardened steel thrust button pressed into the front wall, to control countershaft/clustergear movement and enhance durability of the casting (A356-T6). The 1st gear uses less helix angle than the M22, to increase strength, while still benefiting from helical overlap. The multi-piece ("modular") cluster gear uses tight press fits and huge DIN snaprings to increase the number of ratios (27 and climbing), reduce maintenance costs, allow more modern manufacturing processes (ground gearing) and the "stub shaft" of the cluster is the European equivalent to SAE9310. Our synchro assemblies are lighter and tighter fitting, we use forged synchro rings, bronze-backed thrust washers, a lighter weight mainshaft, a C355 aluminum midplate and a REM-type polishing process to deburr and polish the internals. Even the sidecover and shift forks have gone through design revisions to enhance durability and strength.
Heres a link to one being put together; theres one mistake in the voice over: our sidecover isn't steel, and we now use bronze bushings instead of the stacked caged needles seen in the video.http://www.powerblocktv.com/episode/MC2 ... Bd-w8UhFxV