Contact with both aircraft was lost while they were in cruising flight at their assigned altitudes, which were 2,000ft. in the case of Star Tiger and 18,000ft. for Star Ariel. It follows therefore that the Sperry A3 autopilot was engaged in both cases. If the autopilot is set to maintain heading and altitude, it will make constant adjustments to the control surfaces in order to do so. Imagine a scenario where one of the outer engines gradually begins to lose power. If the aircraft is being flown manually the pilot would feel the tendency to yaw and make corrections initially with the rudder pedals and then by adjusting the rudder trim.

Of course it would be expected that a reduction in engine power would be picked up by fluctuations in the instruments. This cannot be guaranteed however, and in the case of Star Tiger, crew fatigue could have been a factor, to the extent that in the latter stages of the flight it is possible that only one crew member would have been in a position to monitor the instruments at certain times.

Let us suppose that power in the port outer engine is gradually reducing. The aircraft would try to yaw to the left and the autopilot would therefore deflect the rudder gradually to the right to compensate. This situation could exist (and steadily worsen) over some time without the crew being aware of the problem, as the aircraft maintains the selected heading. Star Tiger was close to Bermuda at the time of the last radio contact. It is reasonable to assume that not long after the final radio message was transmitted, the handling pilot would prepare for the approach to Bermuda by disengaging the autopilot. At that moment the autopilot would cease to control the yaw and because of the asymmetric power situation the aircraft would  suddenly and violently roll to the left. It is very possible that this movement would be so severe (and unexpected) that control would be lost. When one considers the cruising altitude of only 2,000ft. it is unlikely that control could be regained in time to avoid a disaster.

Star Ariel was known to be flying at 18,000ft. when the last radio message was received, and not close enough to its destination for the approach to commence. However, the same scenario could easily fit this circumstance too. If a member of the crew noticed a reduction in airspeed (because the aircraft was yawing severely) they may have suspected a problem with the autopilot and disengaged it temporarily to troubleshoot. Again, the aircraft could have turned so violently that it entered a spin which was either difficult to recover or resulted in a structural failure of the tail. Either way it wouldn't take long before the Tudor hit the water, even from 18,000ft.

Of course, all of the above is pure conjecture. But it is surely not beyond the realms of possibility.