That 0.02% figure you are referring to is probably from here
this. The previous day is is 0.20%, not 0.02% for starters which is drastically different, but this only shows gross short sales for that previous day. We only get the aggregated data with T4 delay. That % is based on the product on issue which is another word for shares on issue. What the shares on issue are and what the free float (tradable) shares may be quite different. I track the net change as a % of volume and some days it goes above 20-30% of daily volume and that assumes that short sold shares are only traded once in a trading day. This was occuring a lot in months like October where the SP was slowly decreasing with not much volume.
The other thing that distorts this picture is the fact shares can be traded back, and forth which increases the volume but it doesn't necessarily result in a large net change in position. These institutions don't pay brokerage on transactions and these trades are largely placed by algorithms.
Re volume: For example UBS (as a group) has sold about $63m of stock but has generated close to 1bn of volume. They don't have access to 2/3 of the shares of this company.
(20min delay)