Since this has now come up more than once, maybe it would be worth adding code to delay the initial call.
Apparently, not everyone is affected by this issue. We'd need someone who is experiencing the issue and willing to assist with tests (starting with answering the "does it happen with a different card" question I tried to ask in the other thread).
The M10 port has extra code for delaying spytask until task_Startup finishes, but I added that to prevent a crash.
