My memory from an OEU presentation at Laarbruch in 1986 is that no full operational drops were trialled in that the dispensers were always retained at the end of each drop. This was due concerns about safe separation of the empty dispensers at operational speeds and heights. Thus they were jettisoned at 250 kts in 25 wing on a second pass. Therefore the first full operational trial was carried out on the first night of Desert Storm.
No aircraft were lost dropping JP233 so to call it useless or tactically ineffective is just plainly incorrect. One aircraft was lost post-target but this was mostly likely a CFIT event. There were occurrences of nav computer reversions to a degraded (Doppler SAHR) mode at the end of a JP233 run which had an immediate effect on the accuracy and latency of the HUD symbology. Thus bunting down by night in reaction to the ballooning effect of losing over 5 tonnes of mass from the aircraft over a few seconds was particularly risky.
In terms of weapons effects, I think they were reduced somewhat by the soft sand beneath the Iraqi runways as the heave effect of the secondary warhead relied upon a good dense subsoil to direct the blast.