Which is basically the direction we all have been heading for ..... but the Hughes protocols demand that all else is proven not to be at fault first.
Grass is green
Sky is blue
Its always on the users end.
At that frequency of testing, watch your data consumption.
My guess would be that once per hour would be plenty. Once every tow hours even would be well beyond the 5 tests (for average) 3 times daily as means to quantify an "issue" versus a "prime time" problem.
That's a lot of data going down the drain Bob.