The missed approach point (MAP) needs to be identified somehow by the pilot flying this approach.
The example you show is a VOR approach that requires neither RNAV/RNP capability nor DME. That is also why you see a duration (3 Min or 2 Min) for the outbound leg instead of a fix (there is an optional RNAV fix [FS044] that may be used if you are RNAV capable, but it is not required).
The only point that can be identified here is the VOR itself. Therefore, the procedure is constructed with the VOR as the MAP, even if that happens to be after the runway. Note that you can execute a missed approach earlier than the MAP (that is the latest point, where you have to execute it).
If you look at the similar VOR Z Rwy 04 approach with straight-in minima, you see DME required at the top, which allows defining a MAP before reaching the runway via DME:
