The thing is, remote monitoring like this can be handy as a developer. You can design the best testing procedures and systems on earth, but you can't entirely accurately simulate the impact of 100s of users.
As such, these systems can be a godsend for diagnosing the fault.
It's not quite the same, but I maintain a PHP based website at work that uses a java based webservice to do a lot of it's donkey work. It has a simple fault monitoring system that fires off emails including various bits of diagnostic information (things like user ID, parameters passed in and stack traces etc) if a variety of things occur, such as exceptions being thrown. The result of the system doing this is I can often have the problem at least diagnosed (if not fixed) by the time the user reports it. It's not a full fault reporting system as such (it does not store details of the faults, so the only why I can analyse them is by keeping the emails) but it does the job it was intended for..
I'd be a little wary of these services offering to store this kind of info for free.I've never really looked into how much the paid services charge, but given an app with a reasonably large user base, this app sounds like it could potentially generate a lot of data. Data that would need to be stored. Storage costs, so I'd be wondering where the company that so generously offers to store this data for free is deriving it's income.