The one I have seen that failed in a Stratos replica (former member on here) was much more catastrophic and the weak point was actually around the outer edge of the end casing.
The increased traction of the mid engined layout must have been too much and the thrust put back up the shaft from the crown wheel was too great for the casing.
So this was the sort of failure we were attempting to avoid with the additional strap


Now 'm not saying this is absolutely correct, but here was our thinking on the situation


- the failed section is after the gearbox mount, so it won't be subject to any loads generated between the diff output to the chassis


- the only Force that part of the gearbox casting is subject to is created from the gear shaft bearings


- As the gears are loaded up the shafts will try and spread apart, this is reacted by the bearing housings in the casting


So our thoughts were, a failure occurred in the casting between the 2 bearings, as mentioned it is very very thin in there, once a crack formed the 2 gears spread apart further and caused the carnage seen in that photo


What would be super helpful would be a photo of the failed casting parts, if the 2 bearing housing are still attached together in the end plate that would falsify our theory