According to an answer by John Frazer
The only successful plane with inhabited wing might have been the Ju-38
Is this true?
In the whole history of aviation there was only one aeroplane with inhabited wings (so that people were able to stay in the wings, not just for repair purposes)?
What are the reasons that there were so few of them?
There are a few more aircraft with seats in the wings, but you are right as long as passenger seats are concerned: they appeared only in the G-38. Now I could also mention the Soviet Kalinin K-7, which was supposed to seat some of its planned 120 passengers in its wing, but that design never went beyond early flight test of the first prototype.
Professor Hugo Junkers was convinced from the start that the best aircraft would only consist of the wing and engines, but the size of real-world aircraft did require a fuselage to place pilot and passengers. Only when wing thickness grows to 2 meters would it be feasible to place people inside (the root thickness of the Junkers G-38 was 1.7 meters, that of the K-7 was 2.33 meters). That was only possible with the largest aircraft of their time, like the Messerschmitt 323 which had engineer's positions in the leading edge of the wing.
The Illustrated London News article about the Me-323 from 1943 contained the picture above (source) which indicates the engineer's cabin between the two inner engines of each wing.
The simple reason for the scarcity of inhabited wings is the limited chord and thickness of practical wings. Back in the days of propeller airplanes, the relative wing thickness could be 20% or more, but with modern jet transports, even 14% are on the high side (this text contains an extensive collection of data). Next, the wing of airliners is full of other stuff:
which leave no space for passenger seats. Only in the days of slow aircraft with low wing loadings was the inclusion of passenger seats even conceivable, and only in the largest wings. The G-38 had a wing loading of only 80 kg/m² while the A380 has one of 680 kg/m², so its wing area is only three times larger than that of the G-38 and its maximum root thickness about the same. If you now consider that just 6 out of the 34 passengers on the G-38 could sit in the wings, it should become clear how infeasible it becomes to place passengers in the wing of even the largest aircraft of today.
Junkers went even farther in turning the blended wing into reality when they (together with Messerschmitt) had to submit proposals for a large cargo glider in 1940. While Messerschmitt scaled up their then-largest aircraft, the M-18d, by a factor of three to arrive at the conventional Me-321, Junkers opted for an all-new design which was only loosely based on the G-38 and blended the cargo volume into the center wing. Their Ju-322, however, turned out to be almost uncontrollable. In the haste of the development, nobody cared to check how effective the far too small and too closely coupled tail section was.
Theoretical concepts of aircraft which place their payload in the wings have focused mostly on freighters, like in this NASA study (PDF!).
