I've used several of these lenses, and they are MOST DEFINITELY optimized for high density sensors. There is no question they thoroughly outresolve my Canon 7D, which has a 4.3 micron pixel pitch. I would not be surprised if Canon's new generation of lenses is prepped for at least 60-70mp FF sensors down the road, if not much higher resolution sensors than that.What is the business idea behind something like that? How many years will pass before we see sensors like that, let alone Canon sensors like that?
Making a lense that will still shine in 20 years sounds really cool. But how much does it cost? What is the risk that they'll end up with a nice set of chess pieces that cannot be sold because unforeseen interior/exterior changes (such as the move from FD to EF) pressure them into doing something else?
First, realize that Canon has already done this. It isn't an idea, it's a simple fact. Rent the EF 600mm f/4 L IS II lens, and you'll see what I am talking about. Anyone who thinks a Canon 7D is soft, pop on a 24-70/2.8 II, 300, 400, 500, or 600mm L IS II, and the sheer resolving power of the 7D will finally reveal itself...without any softness anywhere to be seen.
As for the business value behind Canon's new lenses, I figure fundamentally it is the same as what was behind the original versions of the same lenses they replace. Canon's previous 300, 400, 500, and 600mm lenses all lasted for well over a decade, with an interim update to add 2-stop IS. Those puppies upon introduction cost roughly the same as the new generation does, which ranges from ~$8000 on the short end to ~$13,000 on the long end. I don't see any significant nominal price difference, and factoring in inflation, the current generation of lenses is actually cheaper than their predecessors. They are the most-used professional grade lenses in the world, ubiquitously seen at sports events, the Olympics, used by wildlife photographers and bird photographers, pretty much anything that needs the best quality and the greatest reach, at least two out of three times (if not much more than that) you'll see someone choosing a Canon supertelephoto lens. While Canon may no longer be the creme of the crop when it comes to sensors, I don't think you will find many photographers who even had the thought to question the quality of Canon's glass. It is second to none, and with recent advancements, I'd wager that Canon glass is just as food as Zeiss most of the time, if not better in a few cases (such as the 300mm and 600mm Mark II superteles.)
From a technological standpoint, the resolving power of the last generation of Canon lenses WAS beginning to show it's inferiority in light of increasingly dense digital image sensors that are finding their way into cameras. While techy geeks like you and I may understand the value of the sensor outresolving the lens, the majority of photographers expect impeccably sharp images strait out of camera. There is no questing that Canon's new line of lenses outperforms, in many cases by a very significant margin, the lenses they replace. The only logical conclusion I can really come to is that Canon is preparing for a future where the average sensor resolution is higher than today by enough of a margin that it warrants replacing all of the significant lenses in their L-series lens lineup.
As for when we might see sensors that need this level of resolution...not long. The 7D clearly needed better lenses than a significant majority of Canon's lens lineup prior to the release of the Mark II generation (pretty much every lens I've used on it resulted in some softness to some degree, with the exception of the 300mm f/2.8 L IS first generation and the 100mm L macro lens, and probably the 24 TS-E, 400, 500, and 600mm lenses from the same generation.) If the 7D Mark II really does land at 24.1mp, that is equivalent to a 62mp FF sensor...so we might only have to wait a few months before we see the kind of sensor that can thoroughly utilize lenses with much higher resolving power. If we get 24.1mp 1.6x APS-C sensors this year, within 3-4 years I think we could see 32mp APS-C sensors...or the equivalent of 82mp FF sensors. Within a decade? Hell, if we keep increasing sensor resolution by 20-30% per generation, we could see 50mp APS-C sensors and 140mp APS-C sensors within the decade. I don't know how likely a linear progression of sensor density is...but we already have sensors with pixel pitches approaching the 1 micron scale that produce great IQ...so I don't think it is technologically improbable by any means.
Finally, regarding a mount change...I don't foresee that happening any time soon. EF was a calculated move that Canon made, knowing the risks, in order to open up new possibilities that were not possible or too costly to apply to the FD mount. Given the fact that Canon has expanded EF mount compatibility to the EOS-M line, which should have a long and healthy life decades into the future, and given that Canon has demonstrated an ability to enhance their AF and IS capabilities by improving the firmware of lens and camera without changing the mount, I think EF is here to stay long into the future. I have no worries about a mount change.