Guys,
Perhaps I can help here.
The SHAR asymmetric crash off the ramp at Yeovilton was due to failure of one of the two 330 gallon ferry tanks to transfer fuel. Pilot didn't notice and the resultant asymmetry resulted in losing roll control on launch. Possible contributory cause was (I believe) a plethora of confusing and partially embodied Tech Instructions on the tank pipe couplings. Drop tank fuel state indications were also not obvious. We fitted 330s to our squadron's jets in 1989 with no problems once we had put them through the Tank Bay to get the TIs checked and fitted. Usually fitted with the fixed AAR probes at the same time.
DBTW and others: 190 gallon tanks were indeed modified version of the Hunters' 230 gallon tanks. Not welded, but three sections screwed together. They could be jettisoned perfectly happily when full (for example on launch) and also when empty as long as twin Winders weren't fitted outboard, when they could catch a fin. The worst jettison case was half empty when the post jettison path of the tanks depended on where the fuel had sloshed to. In sum, the 190s jettison characteristics were really no worse than any other empty or partially empty drop tank on most fast jets.
We (DGA(N)) were able to put together a perfectly safe RTS for the 190s on the FA2 in 1993.
That said, the 190s were unreliable beasts, mainly due to poorly engineered electrical connections and the internal plumbing. They were a rush job, looked like a rush job and behaved like a rush job. Definitely not one of Kingston's finer products....
Plastic, I'm afraid you've been misinformed - there is no real 'lee' effect behind a ski jump - the wind curls back down over the lip and is doing a fair lick by the time you get around 150 feet aft. Ramp launch calculations don't include any correction for any 'lee' effect. Kingston did a lot of work getting the nose leg loading and extension matched to the ramp, as well as handling the 'unload' extension, as you rightly say. I've not heard of any legs being snapped on UK jumps.
USN did experiment with a concept called 'Sea Launch' where they tried both 'free take off' ramps and also fitting a ramp immediately at the end of a steam cat. As for most CTOL aircraft, ramp benefits were slight and required redesign of the aircraft to improve them.
As JF has pointed out, an ability to adequately control the aircraft at low post launch speeds is essential. As he also inferred, you would need a very (very) high basic T/W ratio to get off a ramp in a CTOL aircraft carrying a max war load. In my view, probably stupidly high. Vectored thrust aircraft are better for ramps because they can use jet lift as well as wing lift to arrest rate of descent after ramp exit, thereby allowing a lower ramp exit speed, and a shorter TO run. They also already have the required low speed control systems fitted. I'm quite sure the 'Sea Typhoon' guys will have worked all this out. Or not.
Best Regards as ever,
Engines