I think this could be a valuable service, but $99/test seems quite high (compared to similar services targeting desktop instead mobile users). At $500 to run Nielsen's recommended 5 person test, running my own tests starts to look like a viable substitute for your service (considering I'm likely to want to run more than one round of tests).
I think this is true at both ends of the scale: Startups will compare this cost against guerrilla testing techniques (e.g., sourcing testers on fiverr, asking people on the street, or inviting them into a lab for a hour for a small reward). Large firms will compare the value of this service to hiring a more high-touch UX firm (or building this competency internally) that does much more than record test sessions.
For what it's worth, here are some questions I have as a potential buyer:
* How long are the test sessions in general?
* How are the tests "facilitated", or more to the point, what happens if testers don't understand the task instructions or can't get the app to run?
* How are you sourcing testers? Can you quantify their demographics? Can you provide testers that meet specific demographic profiles?