then does it mean such short position will never been covered if the user does not add margin?
It means the losses are eaten by the network rather than the short position. The network earns the BitUSD necessary to cover these losses from transaction fees and in most cases accumulates a reasonable buffer prior to the event.
When a short positions runs out of collateral the network simply discards it as no rational actor would bother to pay additional losses.
Of course everyone trading must be aware of the assumptions baked into the blockchain prior to investing:
1) Rapid loss of value of 50% in hours is very rare, even for bitcoin (once the network is established)... can anyone show an example where Bitcoin lost 50% in hours (excluding the Mt. Gox flash crash)?
2) A loss of 50% is still break-even for the network, assuming everyone was fully collateralized before the 50% drop.
3) The most likely scenario for a 50% drop is after a large increase (bubble) which should be less likely given most shorts would use a BTS bubble as a covering opportunity (suppressing the bubble in the first place). The demand for BitUSD will increase as the bubble grows which keeps the bubble in check. Remember, BitUSD is the 'short' position against BTS.
4) The rise in value of BTS means that the collateral backing all existing BitUSD went up in value dramatically and thus a 50% fall is unlikely to even consume all of the collateral except for anyone who shorted BitUSD at the top of the BTS bubble. Few people would want to open a leverage position after such a massive run up and thus the amount of BitUSD that would lack sufficient margin to cover would be relatively small.
5) The worst case 50% fall that results in insufficient collateral would have to occur at a time when there was no recent run up and thus be triggered by something other than a bubble collapse. This would be like Bitcoin losing 50% of its value in an hour after having price stability for months. Possible trigger events include: finding a bug in the blockchain, shutdown of the internet, or some regulatory action. During such an event, perhaps after a long-slow fall in BTS value it is possible for many positions to start out with some loss and thus require less than 50% immediate drop to trigger an insufficient funds scenario.
6) The holder of BitUSD is not protected against systemic risks associated with all crypoto-currencies, only from the volatility that is independent of those systemic risks.
So what is the economic effect of having some shorts blow out and thus leaving more BitUSD in circulation than short positions that need covering? A fall in the value of BitUSD relative to BTS proportional to the unbacked BitUSD in existence. What is the economic effect of having some BitUSD destroyed by transaction fees? A rise in the value of BitUSD relative to BTS proportional to transaction fees destroyed.
However I believe that these supply driven theories of value will only play a part on the fringe / extreme cases and that the global consensus that BitUSD should be traded at a price point near real USD will have a far greater impact on the demand side of the equation.
Lastly I would like to make one last observation, large market moves that result in significant un-collateralized BitUSD will result in one-time devaluation of BitUSD.... but this one time devaluation would simply create a new constant offset from BitUSD and the market will continue to operate with very high correlation to real USD. In other words, BitUSD is useful as an investment vehicle due to its correlation and not its price. What we are pegging is the correlation and not the price.
Thus those long BitUSD have an insurance policy against BTS price drops with a maximum payout of 2x. They bear all risk beyond 2x.