The winglets have to be designed for a given environment. No doubt, the winglets were designed for airline ops and not military ops.
On the other hand, the difference between normal cruise and high speed cruise is often negligible in time but incurs, of course, a significant fuel burn. It seems to me that scheduling is a problem if they require the airplanes to constantly deploy at high speed cruise.
In days of old when I was driving the water wagon KC-135, we cruised around at M 0.80 but when in SEA on refueling sorties, we let it eat.. set max cont thrust and take what it would give you until you hit M 0.90 and then you throttled back. But we are heading to fighters short of fuel or damaged coming off targets and needing fuel NOW.
And again, adding winglets is not just bolting on a stuff. It is redesigning the wing and that always creates problems. You are working at the weakest point of the wing, changing the loading, creating new stresses and new problems.
FWIW, the first tests Whitcomb did with the winglets for NASA were on a KC-135. They worked but the USAF opted to go for engines instead of wing mods.
NASA then bailed a DC-10 from Continental and it too worked but no one wanted to pay for the STC that would have been required.
The first real civil use of winglets was on the Lear Longhorns, the Lear 28 and 29 and they were really created as technology demonstrators but wound up with some buys. I don't think any of the 29s are still flying but as I remember some of the -28s are still on the FAA registry. And if I am not mistaken, they were the first ones to be certified to FL510... a LONG time ago.