Reassembling the action off of the wood can reposition the triggers to the sear arms.
The screws aren't usually tightened the exact same amt as when the action is in wood plus the bottom tang can actually flex/bend a bit when the action is assembled in the wood as opposed to being off of the stock.
All that means that a trigger or triggers can be pressing on a sear arm or arms slightly when assembled (on the stock) and not so when assembled off of the wood.
That slight pressure is the same as slightly pressing the trigger and is disengages the sear from the hammer.
Check with the gun assembled and see if both triggers show a bit of slack in them when the gun is cocked. They should be able to be pulled away from contact with the sear arms and then spring back into contact under the trigger springs pressure.
Solid contact with no slack usually means the sear is being affected and cannot fully engage the hammer sear slot.
Also, the sear slot cut into the hammer can collect dirt/dried oil, etc in the bottom of the slot and keep the sear nose from fully engaging when cocked.
Check sear cuts in the hammers and the sear nose on each for crisp edges and slot. wear and someone trying for a better trigger pull can lessen the engagement.
The mallet slapping you gave the action to see if it would drop a hammer is OK, but it does not equal the recoil of a real live fire.
About all I can think of right off.
