kenglish has got the best suggestion. Instead of having to worry about 9 million customers trying to bring in the money, just have 100 or so. Don't create the programming. Just own the satellite, the broadcast license and the facilities to uplink to the satellite. Money can be made this way - look at SES Astra - they don't do any programming, all they do is lease transponder space to TV and radio stations on their satellites. Plus there's plenty of Sirius & XM reception equipment out there already, so it's not that there has to be a great big marketing push to get new receivers to market.
If Sirius/XM want to retain a paid model on one system and go free-to-air on another then it would have to be worked out carefully. Let's suppose Sirius/Xm are going to abandon the Sirius system and go with the XM setup. There'd have to be massive sweeteners to entice Sirius equipment owners to acquire XM equipment, the two billing systems will have to be merged (right now they work very separately; I received a Sirius "dont leave us" offer, called XM and they said it's strictly with Sirius), and there would have to be time spent in switching over (possibly achieved by signal degradation, channel number reduction, and complete "switch off" with a period of furlough before the activation of a free service.)
If Sirius/XM want to retain a paid model on one system and go free-to-air on another then it would have to be worked out carefully. Let's suppose Sirius/Xm are going to abandon the Sirius system and go with the XM setup. There'd have to be massive sweeteners to entice Sirius equipment owners to acquire XM equipment, the two billing systems will have to be merged (right now they work very separately; I received a Sirius "dont leave us" offer, called XM and they said it's strictly with Sirius), and there would have to be time spent in switching over (possibly achieved by signal degradation, channel number reduction, and complete "switch off" with a period of furlough before the activation of a free service.)