Landing straight ahead into any water is far better than stalling, crashing and burning trying to make it back to the airport.
Compare the outcome of the following two engine failures in PA-32 aircraft:
200204328
vs
200802048
In the first turnback attempted - literally crashed and burned and six people died.
In the second ditching accepted - five? people bobbed around in the ocean for a short period of time before being rescued.
Yes - turnbacks are possible in a high performance aircraft - it is a standard procedure for ASEPTA approved aircraft (IFR charter flights in single engine turbine aircraft) - and ASEPTA pilots are trained in the procedure.
Many people have died attempting turnbacks in piston singles - particularly when the aircraft is loaded - most people try it empty or with just an instructor on board. Whilst the glide angle doesn't change with extra load, the gradient achieved during climb certainly does. Add in high temperatures, high humidity (affects piston engine performance significantly) and high altitudes and the probability of getting back to the field reduces dramatically.
Also consider the energy equation with 15 knots headwind on takeoff and a touchdown speed of 45 knots:
Land straight ahead and the touchdown ground speed will be 30 knots.
Land in the opposite direction and the touchdown ground speed will be 60 knots.
Energy is proportional to the square of velocity - double the velocity equals four times the energy.
I don't know about you, but if the plan doesn't go perfectly I'd much rather have a quarter of the energy at the crash site.
If you are an instructor and recommending turn backs I hope you have excellent professional indemnity insurance - because if someone stuffs it up and the pilot survives and says 'my instructor told me to' the lawyers for the passengers or their estates will be rubbing their hands with glee.