Have you let them do it so that the lesson is indelible or just vetoed the plan in the interests of the conduct of normal operations?
In those few words, you have described a common factor that can crop up when the captain elects to give a first officer a "leg."
Do you let the first officer paint himself into a corner and thus hang himself during the conduct of a descent leading to an instrument approach - if, for example, the aircraft has been allowed to become too high and fast for comfort. Why not let him go so that the "lesson is indelible." After all - goes the argument - a go-around is always available if the approach goes pear shape and cannot be salvaged by final approach stabilisation point. Keeping in mind, this can become a commercial issue with on-time arrivals paramount to the reputation of the airline - not to mention the inevitable media attention from a go-around.
Or, should you say "Taking Over" early in the piece, where you can see problems ahead getting the aircraft legally stable. Never mind hurting the feelings of the first officer - that discussion is better left for the debriefing once in the crew room.
What should not happen, is the captain leaving his decision to assume handling control so late, as to require some fancy footwork on his part to get the aircraft safely on the ground without breaking the rules on stabilisation. Far better to pick the flight path situation early and take over then and there, so there is a seamless transition from first officer to captain well before drastic power/speed/speedbrake/flaps/ are needed by the captain to recover the situation.
An example from a recent approach: 20 knot crosswind with F/O as PF. Let's say it becomes obvious by 1500 ft on final that he is not coping too well and is allowing the aircraft to parallel runway direction slightly downwind instead of grabbing the localiser or runway centreline and tracking accurately to the flare. The captain doesn't like what is happening and urges the PF to get on centreline before it gets too embarrassing. The PF makes abortive stabs at a heading but is not within a bull's roar of getting smack on centreline until almost over the fence. We have all seen this during crosswinds. Some captains will hope upon hope the PF will eventually get on centreline - just give the bloke some slack -and his urgings get more insistent the closer the runway is. The wings are levelled over the fence and the landing is hard with drift still there. The captain curses inwardly for allowing the situation to develop to where either a go-around is necessary or he allows the PF to fight his way onto the runway but no time to get the drift sorted before touch down. Unless of course, the PF is a very good, switched on, pilot.
It is probably a familiar scenario that most captains have experienced in their career. . There is often a reluctance by captains to interfere with the other pilot's decision making or aircraft handling. After all, the first officer has got to get the experience sometime - right? True of course; but arguably not at the expense of the company when a go-around may be the result of poor flying ability. And not at the expense of frightened passengers who expect two highly trained professionals to be in full control up front. We haven't talked about cadets as second in command, so lets not go down that path.
There needs to be a compromise between the captain allowing a certain amount of leeway to the first officer to get on with the job of flying the aircraft on his leg. And on the other hand, deciding at what point in an approach the captain should take over to complete a safe arrival if errors of judgement by the first officer become too much to accept. Most would argue it is better for the captain to err on the safe side and take over early -rather than too late to avoid last minute fancy manoeuvering?
Points to ponder, maybe?