I would assume free cash flow would be the reason. Let's assume that you could get $20M per seat on a 7 seat dragon with 2 of those seats going to crew and the other 5 going to paying passengers. You, launch them on a re-usable F9 which you currently charge $80M for a non-reusable booster flight into low orbit. Could you clear 25 - 35 million in margin on that? Could you do that for doing "nothing" special other than having the gear ready? Clearly we'd need way more visibility into their operational costs but if you could generate 250 - 300M$ per year that was not from fixed price contracts I think that would be a helpful thing from a money management perspective.

They still might not do it, but I don't think the sub-orbital passenger trips are very close. So many "other people" have to get involved before that is real, like treaties with the world about how you regulate overflights, or tariffs, or do air traffic control, or who pays when your rocket augurs into the innocent village, Etc. There are lots of details there, whereas Roscosmos has already done tourist flights to the ISS, and with their own launch facilities most of the variables would be under SpaceX's ability to control.