Jano, you remember correctly. We both read this comment in the Erwin Puts review. I was curious about this & so I wrote to both Mr. Puts & to Carl Zeiss AG. I received replies from both of them.
Erwin clarified his statements in the review, which said at one point that the M3 rangefinder has been extremely durable & at another point that it could be easily knocked out of alignment, as follows: "There seems to be a contradiction in my remarks: the M3 finder is a one piece design & does not need adjustment as there is nothing that can be dislocated. But the M3 one piece construction is located very close to the top cover and frequent crashes on the top cover by, as example, a second Leica body or by abuse as the camera is laying on the floor in a truck, can dislodge the finder assembly." He contrasts this with the rangefinder in the M6, which was designed for the M2 & is found all successive Leica M's: "The M6 finder series is a two piece construction and sometimes needs some additional adjustment."
So, what I read Erwin to be saying is that the finder in the M2 - M7 will need periodic adjustment as the result of normal use and the finder in the M3 will not. However, the finder in the M3 is capable of being knocked out of alignment if subjected to repeated ("frequent") abuse due to its interface with the top cover. Thus, the vulnerability of the M3 finder has nothing to do with the design of the finder but with the overall design of the camera & the way its parts fit together.
From Carl Zeiss AG, I received the following reply: "The optical path of the M3 rangefinder was more complex than that of later Leica models It contained 3 prisms for beam deflection while all later Leica models have just one. Such a more complex system is more elaborate to adjust during assembly and at least theoretically the risk of later misalignment by later mechanical stress on the camera is higher. But a theoretical problem is more an academic issue if things are well made. Not only the long product life of the M3 from 1954 to 1967 suggests that it was a reliable piece, but it was in the 60's that a Nikon F & a Leica M3 were the regular companions of photojournalists going to rough places."
So, like Puts, Zeiss stresses that a "well made" rangefinder assembly should not be a problem although misalignment is theoretically possible. They note that it is during assembly that it is vulnerable to alignment problems because of the complexities of the design. It seems to me that Zeiss quality control, which includes an inspection of every camera before it leaves the factory, should catch anything as obvious as rangefinder misalignment. So, having a ZI end up in a customer's hands due to improper adjustment during assembly is highly unlikely.
I should note that the ZI rangefinder design is not a replica of the M3 design; there are differences. This is obvious in use in that the rangefinder patch remains stationary in the M3 while it moves as focus is adjusted in the M3. It seems to me that the ZI design is more an elaboration of the design developed for the CL, made more complex by the requirements of a longer base line. What the ZI design has in common with the M3 is that unlike the M2 - M7 design, the frame mask is parallel to & close to the illumination window. The result is increased brightness, reduced flare, & elimination or rangefinder parallax.
Huck
