It’s fallacious to equate volume with demand. For example, if I buy a BTC for $240, sell it for $260, buy it again for $255, then sell for $245, all in 24 hours, I created $1000 of volume with my money, even though it’s impossible to buy $1000 of NBT with that money.
You can’t lose all by losing the peg to the upside. “Oh, no, my NBT are worth too much! I guess I’ll… sell them? I guess I won’t buy more?” And the problem fixes itself, because the buy walls at $1 remain. You can lose some reputation and volume in this way, but the system returns to $1 eventually, not to $0 as could happen in the case of losing the peg to the downside. Custodial buy walls are strongest when the peg breaks out upwards, so that’s the moment when the risk to “lose all” is least.
This could even bring more publicity to NuBits, if spun the right way. Imagine the headlines: “NuBits fails by being too successful!” 
Again, I’d be interested in custodial data on this. Has anyone been collecting getliquidityinfo over time, so we can see a graph of custodial liquidity and what size of relative shock would be needed deplete our reserves?
Also, as several people have mentioned, would Jordan be willing to make public the total number of multi-sig key holders?