1) give or take yes
2) No, if they short #'s = 50% they were *at most* able to cover all the shorts they opened for that day only, there would have been no reduction or no addition to long term shorts. But the *at most* is important, they could have decided not to cover any of it so it could have been 50% more of the day's volume added to the permanent short pile. Truth is unknowable but likely somewhere in the middle. If the % is >50% for the day, let's say 54% like on 8/4, then *at most* they could have covered 50% of the day's volume, leaving 4% of the day's volume added to the permanent short list. But they could have added more if they didn't cover as much. If the % is <50% for the day, say 45% as on 7/29 then they could have covered all the shorts they opened for that day plus another 5% of the day's volume in longer term shorts. Again, the likelihood that they sit at the extremes of these numbers is not good, somewhere in the middle more reasonable.
3) No, it doesn't tell us much until you start tracking it over multiple days. Near 50% on the day is too big of a variance in possibility to say much definitive. If the day is very low or very high then you can say something a little bit more definitive about that day's action but there's still a gray area.
Keep in mind that we know the float is locked so there is no way they can be reducing any long term FTD's they may have. That lends the above numbers some more meaning but given that trades can have multiple legs, are 'media reported' and they often don't report more than one leg of the trade, take the daily numbers with a grain of salt.
News 
Market Data 
Discover 
