There are at least two or three things things wrong with that paper, one of which the author admits outright - that he assumes demand stays the same. One of the aims of the aggressive short seller is to dispirit longs and make the stock look unattractive new wild be longs. The result is a very long bow in the demand curve.

Furthermore, You cannot reach a point of "naked" shorting until all shares available to locate and borrow have been shorted. So you don't have naked shorting vs regular shorting, you just have what amounts to aggressive short selling regardless of the ability to locate. Once the shorts run out of shares to borrow, they end up naked shorting. A corallary to this is that the short seller cannot determine when a stock is naked shorted. They just short. The brokerages take the order to short (or the computer executes based on certain assumptions) and a few days later they realize they cannot deliver the shorted shares to the party they sold them to.

As for the shares 'bouncing back' I think the writer is too optimistic. In some cases yes the average price evens out and someone who buys all along or holds through it all evens out. But that is not the real world. In the real world the short is actively using as he sells, perhaps being the primary price driver from the bottom p to the middle p. and yes stocks with binary events could wipe the short out. But the aggressive short dreams of destroying a company be step at a time. He destroys the stock price and it's ability to raise capital, and then it destroys it's reputation. He can't always win but he may win more than he loses, and probably has plenty of strategies to cut the losses along the way.