The likely answer is time between launches. It is likely, though I don’t have the data, that it takes far less time to refurbish and relaunch using the recovered booster then the amount of time it would take to fabricate a new one from scratch.
As such with limited capital you can earn more over a shorter period of time through added launches due to the shorter turn around time between launches. So the second payload still gets launched, just at a different time. Now you’ve made both of the 100M launches and also saved 40M in booster cost. In addition, the two payloads didn’t need to make any compromises between payloads for location requests, altitudes, weight tradeoffs, etc.
I would love to see a business plan for reusable boosters. I have been searching but havn’t seen one. Commercial launch companies like Boeing, United, Lockheed, etc. have been launching satellites for decades and I am sure would have already developed reusable vehicles if they could make a profit doing it.