You did the math, but used the wrong base. 0.58 as a binary floating point number is slightly less than the decimal value:

![enter image description here][1]

Apparently they always round *down* the computed result to two decimal digits (i.e. to the next integer percentage), probably to prevent this page from showing an incorrect "100%" when you're at e.g. `19999/20000`.

  [1]: https://i.sstatic.net/IltGQ.png