We didn't miss the (arbitrary) release target by 50 per cent, we missed it by 10,000ft. It's a metres/feet thing. QinetiQ's Tim D'Oyly pointed out that we hadn't taken into account the air contained between the aluminium oxide grit particles, which accounts for the slightly early drop.
We could have added a tad more grit and rerun, but decided that we were happy with the thing the way it is. It's a matter of suck it and see on the day, but it will certainly work, which is the most important thing.
Regarding the other objections to the mechanism, we don't need a spring or a shorter device, because we're going to use the distance of travel of the plunger, and therefore the total length of the main payload box, to our advantage, You'll see why in due course.
The rubber oxygen tube will not freeze, or come anywhere close to it, since the whole mechanism will be heavily insulated.
Finally, we've been over the idea of an electronic release mechanism before. In a word: pah. There's just not enough shed boffinry involved, and its bound to go titsup anyway.