First of all a little history about the 50/1.2, 50/1.0, and 85/1.2 designs from Canon.
The EF 50/1.2L is descended from the FD 50/1.2 manual focus lenses, and from a optical design standpoint, it has more in common with the even older FL 55/1.2L and its contemporaries (like the well-known Minolta Rokkor 58/1.2 and the much more obscure Yashica ML 55/1.2) than it has with the EF 50/1.0L. But the EF 50/1.2L is also different: it gives up some center sharpness in exchange for smoother background bokeh compared to the old manual focus lenses.
I speculate that this design choice was motivated by a perceived need to fill the void left by the discontinuation of the EF 50/1.0L, but this is not something anyone not working in lens development at Canon at the time can substantiate.
As for the EF 50/1.0L, optically, this lens was quite novel in design. The use of high refractive index glass to correct sagittal flare and aspherical elements to reduce spherical aberration wide open was not a new idea in itself, but the particular implementation was distinctive. The design is not without its flaws: it wasn't especially sharp; contrast suffered, especially in the image periphery. It behaves a bit like the 85L shot wide open in that there's sharpness at some wavelengths and not at others; and sharpness at certain spatial frequencies and not at others. But in a sense, that's what gave the images a distinctive look (apart from being f/1.0), and a character that lends itself to things like low-contrast, low-light portraiture. And of course, like the 85L, it was hard to nail focus and unwieldy. But for a design in a time when film was the photographic medium of choice, you can't really argue that the EF 50/1.0L didn't fulfill any specific purpose. It's just that the drawbacks and the cost (both to produce as well as buy), not to mention its niche nature, made for poor marketability.
As for the EF 85/1.2L design (both I and II, as the optical formula is identical), this is very obviously the EF descendant of the earlier FD 85/1.2L. The formula was tweaked, but the overall imaging performance is remarkably similar. The EF 85/1.2L and 50/1.0L are the only lenses designed to take full advantage of the diameter of the EF mount--they remain the only lenses ever made by Canon for which the rear element glass goes right up to the end of the lens, and goes across the full diameter. No other autofocusing SLR system is capable of such designs.
What is the future? Since the design of the EF 50/1.0L, a number of advances have been made in computational optics, materials engineering, and production engineering. The potential absolutely does exist to create a lens that is optically superior in terms of aberration correction. But Canon is headed in the opposite direction: they have consistently shown more interest in sacrificing fast apertures in favor of IS and using sensor ISO to compensate, because the latter is more cost-effective and leads to lighter lenses with larger design tolerances. The demand for ever-higher resolution sensors (the meapixel race) has further shifted lens design philosophy toward high MTFs.
So, the ostensible obsolescence of the 50/1.0L is mourned by a small but devoted group of photographers for whom sharpness is decidedly not the end-all and be-all of image-making. And with the increasing popularity of DSLR video, one might have a faint glimmer of hope that fast apertures will once again find a purpose. But I don't believe Canon has put any priority on fast primes. I could be wrong. I hope I'm wrong. It would be wonderful to see Canon release a worthy successor to the 50/1.2L or 50/1.0L. But the historical trend doesn't furnish any evidence for that. If anything, we might see some weird 50/1.4L priced around $900, the disappearance of the 50/1.4, and an upgrade to the 50/1.8 II at the old 50/1.4 price point--"and let the $100 bargain hunters buy the Yongnuo," they might think.
Or if this rumored lens is an EF 35/1.4L II, I imagine it'll be upwards of $1600, easily. It isn't that my pessimism is a criticism of Canon--if anything, they are simply responding to the market, which has been asking for nothing but "Sharp! Sharp! Sharp!!!" It's just what I see from the historical trend in the past 3-4 years. A 5Ds-R might even expose flaws in the beloved EF 135/2L.
