Many arguments have been made both for and against use of autothrottle (A’T) in manual flight. In most cases tipping point is aircraft dependent and thus the manufacturer’s advice should be heeded.
Without such advice there are some general issues which IMHO point towards not using autothrottle.
Some installations tend to destabilise the aircraft in pitch (often overcome with experience, but beware unfamiliar conditions). Some older aircraft have a weak or no autopilot–A’T interface. IIRC older 737’s have no control mode interconnect, but the 1011 (for its age) has a good interface, i.e. height control was a function of combined pitch and speed input and not pitch then speed.
More recently auto-flight systems are optimised for combined use, thus manual flight with A’T must coordinate control requirements – akin to normal flight control. However, with excessive use of A’T and autopilot, this may be the only baseline for the pilot’s experience. Thus attempting to mimic the autopilot during manual flight might not lower workload or give consistent performance due to the differences between the autos and the human in the way they use sensors and inputs, e.g. use of rad alt, gain change with speed / configuration, and trim rate.
Other problems arise due to complacency in not scanning speed or mode change, e.g. a low approach aiming for a ‘short’ touchdown (A’T use to control speed accurately). The approach may drift slightly low or encounter a rising ground late in the approach, either triggering the A’T to retard. If unnoticed, and with correcting nose up pitch further reducing speed, the aircraft could be at risk to a nose-high, slow speed, and possibly hard arrival, even before the runway.
IMHO, use of autothrottle in manual flight is an indication of a more serious problem that pilots may ‘loose’ the ability to fly ‘stick and throttle’. This is not so much the physical skills which are taught from basic flying training, but an individual’s confidence in the ability to fly manually in all conditions; note the calls for more manual flight, but only in good conditions. This apparent lack of confidence could spread to the majority of flying; always use the auto-pilot. Also to other systems, leading to automation dependency – lack of confidence in one’s ability to cross check with simple calculations, rules of thumb, etc; always believe the computer – part of the SOP culture?
The industry might have to choose between using experienced pilots to ‘fly’ aircraft (the old view) and pilots who monitor technology and automation (a future view).
There are pitfalls and advantages in both extremes. A concern is that the future could be imposed on pilots with aircraft design, yet most pilots still want to choose.
I fear that any future choice may only be in the title of the job description, not a choice of how or when to use the A’T.