I am surprised at how cheap the 50mm f2.8 was compared to the other lenses, even the 50mm f1.8. Although the f2.8 lens is probably a Tessar derivative and hence of simple design and so should be cheaper, the magnitude of the difference is a little surprising.
It may however partly be a marketing thing - something camera companies have done for years. I recall for example that the Pentax company, before they made the Spotmatic made a range of SLR cameras without meters. They distinguished these cameras by making one early model with 1/1000th of a second shutter speed. I think it was the S1. And were able to put a higher price on it.
Later after having capitalized on this model in the market they realized that there was a big untapped mass market for cheaper models. They also realized that it would be cheaper to simply sell the S1 with a new top plate rebadging it as an S1a and marked with a top speed of 1/500th of a second. The secret though was that there was no difference in the internals of the lower priced camera which still retained the 1/1000th speed (but this simply was not marked on the speed dial). It was still accessible by those who knew this trick, by setting the shutter speed dial one click beyond what was shown as the maximum speed on the camera. And they charged much much less for this camera which was sold as a consumer model alongside the S1 which was their "prosumer" model.
I wonder then if Canon made a similar marketing decision with their 50mm f2.8 as there would have been sunk costs associated with the higher priced lenses and they could, for example, capitalize on the tooling, machining and production line set up to manufacturer the big brother lenses allowing this one to be sold much more cheaply than otherwise might be possible to an entirely different market segment.
Any ideas?