I have done a fair bit of interpreting engineering drawings from the '30s. From that standpoint, and a general engineering view, I have a few observations. They are of course only speculation in the end, unfortunately.
Part and drawing number assignment is somewhat specific to an individual company. However, some generalizations can be made:
1) When something has a part number, that doesn't necessarily mean that it is a single piece-part. Numbers can be assigned to an assembly which in turn then has its own sub-parts, sub-sub-parts, etc. A drawing will also have a name. Each of those have subs which also all have part numbers. When you get all the way down to piece-parts, drawings will typically have instructions on how to make that part. That piece-part drawing is what would be sent to a shop for manufacture.
2) When something has a part number, not only can it be made up of multiple parts, it doesn't necessarily have to be delivered as one piece. The number could refer to a 'kit.'
3) The bigger the company the more anal it is (and really, has to be). The more people there are, and the more complicated the things they build are, the more that communication has to be very formalized. This can mean something as mundane as the name of a part could have to go through a management change approval.
So, the more speculative parts.
One piece vs two
-The fender could well have always been multiple pieces but we just can't see the split. Something like a landing gear fender would be more likely than average to be damaged and need replacement. So it would/could have been a 'kit' where the pieces needed to replace the fender come together (per #2 above)
-Also, imagine removal and replacement of a fender which has been mangled. Removal and replacement of a part like the fender could be difficult to do if it were one piece. You might have to cut away parts to get enough clearance to bring it through the yolk. Two pieces come apart and back together easily. I very much suspect that what we are seeing is two implementations of two-piece fenders.
Possible reasons for the change between fender and guard:
-Since the fender clearly did change, they very easily could have been designed by two different engineers. One called it a guard and the other called it a fender!
-Per #3 above, someone could have messed up and wrote down the wrong word. It then took from November '34 to January '35 to get the change from one word to another approved!
Cheers,
Scott