Wow! That is a failure I've never seen or heard of before. There very likely was no assembly error. The locating pin is likely still in place, and the actual bearing looks intact and complete. I'm wondering why there was a side load pushing that loose part out. There are witness marks on the inside of the broken-off piece.
Could a mis-shimmed output shaft be the cause? Or did someone hammer on the countershaft sprocket end of the shaft for some reason? If the locating half-ring around the large countershaft output bearing was missing, it could have allowed excessive endplay.
In all honesty, I think it could run forever like this, without that end plate of the bearing race/case. Or a machined aluminum plug just pressed in place so oil wells up and feeds the bearing.