As a flight simmer, one of the available weather engines allows historical dates up to 6 years. While there is a lot of limitations (Like you need to connect to the internet), it remains a lot better than the current implementation.
As a flight sim weather engine programmer, METAR and other weather data coverage is enough to inspire weather scenarios for a sim, the rest is made up. Especially close to the ground and away from the reporting station,ground effects can change the low altitude winds to almost anything, cloud cover reports usually don't talk about cloud type, small-scale phenomena like shear layers are missing,...
The high altitude wind patterns are more stable and generally captured reasonably well by the computational models, so they'd sort of work.
The question is - what do you want this for really?
If you're interested in landing a spacecraft in difficult conditions, you really need a flightsim-type weather engine. If you're interested in trajectory deviations during launch and entry, the high altitude layers might be sufficient (these deviations are small though - you need to actually take data and compare to see them, any AP handles them easily).
Again: Would it be possible to land a spacecraft in any weather?
Since I happen to have a detailed flight dynamics mode of a spacecraft and a weather engine, I did the test quite some time ago - no, it is not.
Low visibility you can generally deal with if the navigation system works and the HUD provides the correct runway overlay.
The Shuttle has such a high approach speed that the (substantial) windshear during the final descent as well as even a strong crosswind are 'easy' to correct if you're used to do this in a single prop plane (basically the scaling factor how you 'feel' the crosswind goes with the approach speed ratio, so a 30 kt crosswind for the Shuttle would be like a 5 kt crosswind in a C-172p - and anyone but the absolute novice pilot can handle this.
What's really killing the approach in a T-storm is turbulence though - it constantly throws you off track and it's very difficult to correct for that. I've tried a dozen times or so and failed everyone. I believe GinGin also has tried to land in a T-storm and not succeeded.
So if the interest here is to make the landing experience more interesting, you don't need the average winds, you need the fluctuation around it (aka a good model of turbulence and its interaction with the flight dynamics...)
---------- Post added at 06:16 AM ---------- Previous post was at 06:08 AM ----------
But the weather vane effect should be less since you are towed at the nose...
Also I think its also different because you are no prop and you have no nosewheel steering. Alone the gyroscopic effect of the propeller is a huge force.
It's more the airstream of the propeller hitting the fuselage and wings asymmetrically that causes the rolling and yawing moments (to feel the direct gyro effect of a prop spinning up, you need a WW-II warbird...) but yeah - the situations are hardly comparable, a single prop plane acts rather differently.
A crosswind in a towed glider means you fly at an offset behind the tow plane while it is on the ground - wind pushes you sideward, rope pulls the nose towards the tow plane, you step onto the pedals to fly straight and that's more or less it.
Usually crosswinds are most of an issue during approach and landing, during takeoff it's a transient issue, followed by trajectory control (can't be blown into the airspace of the runway beside you...)
