Don't poll pkpass updates in a continuous loop on errors

Instead, keep a minimum time between two poll attempts. E.g. relevant
when behind a captive portal.

This could additionally be improved by a network status check in
LiveDataManager.