I think the logic is that if a fault could befall one method of locking - if both methods are the same, the fault could theoretically befall both at the same time, whereas different methods can't fail in the same way so the likelihood that one cause could trigger both to fail is far less likely, even though the likelihood of either occurring is infinitesimally small.
ie: if both hydraulic locking cylinders have a manufacturing defect, theoretically both cylinders could fail at the same time. Further - it's not technically possible to detect that one cylinder has failed during normal operation - so an operator wouldn't know one had failed until the other failed too.
This is the main reason why we have coasters where they tell you not to do up your seatbelt, so the operator can perform a push-pull test on the cylinder before attaching the SRS seatbelt.