Note that had it had people on board the pilot would have probably tapped the thrusters anyway
Diagnosing a problem based on a limited number of symptoms available by telemetry is pretty tricky.
There are various possible scenarios that caused this. The obvious one is some kind of contamination in the He line but there could be ones based on Dragons exact thermal environment EG this is happening in Earths shadow, it was cold and this changed some hardware properties.
It does serve as a reminder of what John Carnack has called troubles with "Things with springs," which have always been problematical and all real space vehicles have lots of these components on board.
As for "consistency" Spacex have flown 2 missions which have seen serious thruster use. This is the 2nd.
Spacex have something like 12 cargo missions before they are even thinking about carrying crew and that should resolve of the questions about wheather a difference between their simulations and real flight is significant.
Note that Cygnus has not taken off at this point, won't be upgradeable to crew carriage and any other winner of the crew carriage contract will have to begin their reliability record from scratch, at which point Spacex will have 12 f9/Dragon combos on its list (plus however many F9 launches take off for other customers).
It will be good to finally see Antares launch, even without a dummy Cygnus on the top.