Cornish Jack the Trarom A310 accident in Bucharest does not clearly support the argument for guarding thrust levers. In the accident one of the thrust levers stuck at TO thrust whilst the other retarded; what the crew failed to do was monitor the system / aircraft and disconnect the automation with a deteriorating situation. Thus as you state never assume always check, but checking does not necessarily require physical hands on.
If you imply that guarding controls is lightly resting the hands on the controls / thrust levers, then there will be improved awareness from positional feedback, a form of monitoring (N/A A320 onwards). However, caution for those operators who have autopilots with very light overpower forces (no auto cut-out) as a heavy hand may restrict the auto control input, which could cause the trim to run with more serious consequences than would be caused by a random autopilot disconnect.
I am still puzzled by the number of SOPs that appear to originate from “that’s a good idea”; without any real understanding of the design / certification principles. How many SOPs are backed up with a written explanation of ‘why’ and thus justify the management of perceived risk against higher workload or increased complexity.
Why would discretion’s company use m.s.a as a start monitoring altitude; good altitude awareness (that’s a good idea), but not entirely ‘Standard’ as in ‘SOP’. If your aircraft requires that you guard the controls so diligently, then why not hand fly the aircraft; oh, you are tired, what happens on the sixth sector when the autos are on the MEL, will you really say ‘no thankyou’. As a professional pilot the operational authorities require (or falsely believe) that you and your crew are competent to fly within prescribed limits; what is ‘all over the sky’ to one pilot may be within acceptable safety limits to the industry, but to an inexperienced pilot, or you, it could be a valuable opportunity to improve. Not blaming you – just suggesting that all of us could occasionally rethink the safety strategy.