SSSETEOWF
Describing someone as a liar is generally considered a personal attack, and indeed a professional one. But since Bill Sweetman hasn't sent for lawyers, guns and money to track you down, he must be above it.
At a basic level, the CONOPS based on a 3000-6000 foot runway makes a certain kind of sense. It's also the basis for things like C-17s, in that there are lot of runways around the world that can't handle US fast jets.
However, having been in a lot of meetings when pre-JSF ideas were being hashed around, I can make a couple of observations. One was that the runway length goal was under 2000 feet - I remember one Marine guy saying "if there isn't 1500 feet of straight road in the country I question why we should be there."
Another was that (as happens all too often) the runway quality was usually overlooked. I don't have access to the JSF KPPs in detail, but I don't remember that issue coming up publicly in that context. A 3000-foot strip in Filthistan is not likely to be made of ICAO-spec, 747-proof concrete. Yes, you can reduce the point loads with a short landing, but you've also got a somewhat larger area that may need reinforcing.
At sea: You gain 10 feet of draft relative to a CVN, and perhaps a ship guy (paging Mr Boffin to the thread, please) could tell us how big of a deal that is. You gain on the swings of less training but lose on the roundabouts of maintenance, with a lot of mechanical bits - and if that was the goal I'd take the B-model billions and spend them on CV autoland.
Indeed, when the CONOPS starts to involve 3000-foot strips, 800-foot, 50000-ton ships and SRVLs, and your primary mission is CAS, you do wonder why you need an F-22-priced supersonic stealth jet capable of vertical landing, or whether a rebooted A2D-1 Skyshark would do the job.