Decided to build a Frej scaled down to 90% (15' long/19" beam). She loves the boat and it fits her like a glove. It is much quicker and more maneuverable than her 16LT.

IMO, Mark's idea/concept is an excellent one. And his implementation of the idea is beautiful.
The 'kayak blows around in the wind' problem is a common one. Small and light paddlers are the main victims, but even with larger paddlers, an 'expedition' boat on a day paddle can be a chore.
Reducing the windage (lowering the deck, picking a design without upswept ends) can help, but the best remedy IMO is a boat with more 'immersion' - i.e. less designed displacement - so that the boat is 'planted' in the water more, even with a light load.

That's why the idea of reducing the beam is a good one.

Bill's point about waterline length is important. At 'normal cruising speeds' in my paddling group (3-4 knots) I never have any trouble 'keeping up' in my Mariner Coaster. The Coaster is a 13'5" long boat with a 12'7" waterline length with a 150# load.
http://www.marinerkayaks.com/mkhtml/Coasterw.html

The 'right' waterline length and a low wetted surface make a boat efficient.

The Frej, with its plumb stern has a waterline length that's a good percent of the LOA. I think that the lines picture shows the boat loaded to the design displacement; in my experience there's a bit more bow out of the water in calm conditions. Still, it has a long waterline (desirable) compared to LOA
http://www.thomassondesign.com/en/thumb/700x0/graphics/frej-lines.png
BTW, I noticed recently that Greg Barton states on the 16X and 18X design description that they did testing and found 'no advantage' with overhanging ends. But those boats do have very full (buoyant) bows, like the Coaster and to a lesser extent, the Frej.