Even if you lucky enough to fly something with a TOGA switch you always follow the throttles with your hand.
On B737, dual channel approach with AT, a GA is made by pressing TOGA. The autopilot rotates the a/c to GA attitude and AT advances thrust. IMHO PF should be following through on both controls with both hands during the approach and should do so during the initiation of the GA. PM should also confirm that 'Thrust is set' before gear is elected up. So PF follows through on advancing TL's and PM visually confirms the engine gauges. The TL's are NOT PUSHED. The crew confirm by tactile feedback and visual checking that the automatics have actioned what you selected. It is NOT ASSUMED they have done so.
On a single channel approach with AT the thrust levers advance with TOGA and they are 'followed through' by PF and thrust confirmed visually by PM; same as dual channel.
In a manual approach thrust is manual and GA thrust is set manually and confirmed.
The common feature of these 3 approaches is PF has their hand on TL's during thrust application. Once thrust is set they can choose to put both hands on control column or not. Thrust application is confirmed by PM visually and in all 3 cases PM feels the thrust levers advance. 2 pilots, 2 senses in use.
By just pressing TOGA and PF removing their hand, thrust confirmation is reduced to only 1 pilot 1 sense, visual, and by PNF. Is that the safest method?
What is being said about motor skills, muscle memory, training instincts etc. asks the question: if the normal action is to press & forget TOGA for a GA, but in this high stress scenario it is not the correct action, is the system design to the optimum? Are the SOP's optimum? If the system stays as it is, with the trap, then the SOP, perhaps, needs to be written to include PF having tactile feedback from the TL's; i.e. follow through after TOGA so that if they don't advance they need to be PUSHED. If the SOP is to press & forget then this trap will always be armed ready to bite someone else.
IMHO I do think the system design and the inhibitions need to be reviewed: not only for this Gotcha, but the SFX scenario as well. It has been stated by the consciencous training captains that they discuss/demo/train the traps & Gotchas. i.e. not ALL pilots benefit from this sensible training.
Is it correct that, in this day of more & more automatic dependancy, there should be systems certified with latent sneaky traps & Gotchas that are not documented/highlighted in FCOMS? First the info should be prominent, 2nd-ly the system should be reviewed and redesigned.