Well holy moly! Thanks for making me see things much clearer. Here I thought I wanted top-notch IQ glass because of the IQ itself, sharp corners, client demand and the ability to produce huge prints.
You're welcome :-p ... and to make myself clearer: I think the Canon lineup needs a top 14-24, and the "top of the line" 16-35L could also receive an update for "across the frame" sharpness like the 24-70 did if market demand is there.
However, you were talking of the 17-40L which is a lens placed in the *middle* of the lineup, it's a landscapeish lens so you'll use it stopped down, and it works fine this way. Just as the softness of the 50/1.2 this is not a bug, but a rather feature because it keeps the bulk, weight and - yes - price down. If you want iq beyond that, get a prime (ts), or hope for an updated 16-35L.
But demanding an update of *both* the 16-35L to mk3 and 17-40L doesn't make much sense to me, as a much more expensive 17-50L/4 would catapult it out of the current market position, essentially not "updating" it but replacing it with another lens.
This is the reason why I think requiring every lens from the 50/1.8 to the 200-400L to aspire for top notch iq no matter the cost is a bit on the nerdy side, not to offend you, though your former (now deleted) post didn't really invite a matter of fact discussion I'm also afraid so say.
And yeah, they better quit doing these snobby top-notch IQ upgrades. Or else the price tags of the current lenses will surely skyrocket to untouchable levels! How on earth would anyone be able to afford the 17-40L if Canon released a new version?
Indeed, here you are correct - the 24-70 mk1's skyrocketed to absurd levels after the mk2 release.