I see 3-4 markets:
1. Budget: 17-40 f/4L (this may be a temporary category until Canon goes through leftover stock, though).
2. Landscape: 16-35 f/4L IS - Sharp and nicely corrected, but can only do f/4.
3. Event/PJ: 16-35 f/2.8L II - Not the sharpest in the corners, but f/2.8 allows for faster shutter speeds in events and better subject separation at 35mm. Since this type of photographer would be mainly shooting people, corners are less important as you can't place your subject there anyway due to distortion.
4. Extreme performance: 14-24 f/2.8L - Does not exist yet, but I think something like this will come out from Canon soon to counter Nikon's 14-24. This category is for those who want the fastest and widest zoom regardless of weight and front element size. Probably not the most practical for long events and hiking trips, but to be used when you want to be using the fastest and widest rectilinear zoom even if it will be bulky, heavy, and potentially require an expensive filter system.
I'd rather see a 16-35 f/2.8 III, 16-35 f/4 IS and a 14L III. If the 16-35 f/2.8 III is as good or better than the 16-35 f/4 IS and is able to use front filters, then I think it'll do well in the market. With a class leading 16-35 f/2.8 III, there is much less need for a 14-24, and the 14L III will still be more compact/lighter than a 14-24. Fix the coma characteristics of the 24L II, and you'll great astro lenses at 14 and 24mm.
I agree that some will still need/want the f/2.8 aperture, but no matter how you look at it, the market for the 16-35 f/2.8 II is now a lot smaller if the 16-35 f/4 IS is as good as the MTFs predict. The 16-35 f/4 IS is priced to compete against Nikon's offering. I just sold my 16-35 f/2.8 II, and I really don't care that the 16-35 f/4 IS has IS. My 16-35 f/2.8 II is rarely used because I have primes in the range it covers. What I do care about is its predicted IQ upgrade across the frame.