Whatever the reason, I personally would like to know what it was. This is a very unusual design that at first glance can not be justified by fundamental engineering principles. Although not an automotive engineer, I have a mechanical engineering background and for the last 30 yrs or so have worked with vibration related / rotating machinery problems in industrial applications - many of which were direclty related to OEM design issues.
I concur with JohnZ that there must be a reason why this was done. As John stated, creating this configuration created unique part numbers, inventory, and costs. The key is understanding what Chevy thought was "unique" to the 1st Gen Camaro that resulted in this design. To me it would be a mistake to simply accept this driveshaft design because GM/Chevrolet (the OEM) designed it that way back in 1966 for the '67 Camaro. Not saying that Chevy didn't act in good faith - just that with the additional insight of the past 39 yrs it is appropriate to ask why? ..... why not apply this design to the TH400? ....... why dismiss this design configuration for the 1970 model year?
Someone in Chevy Engineering may have thought he had a better idea of how to eliminate a potential problem with the new Camaro for '67. Was there a resonance problem with the 1st Gen Camaro driveshafts? Both torsional as well as lateral natural frequencies (resonances) are functions of driveshaft material, length, thickness, etc. Did someone try to correct a resonance problem by stiffening the driveshaft with offset yokes? Off setting the yokes would tend to stiffen the driveshaft since it wouldn't be able to articulate properly. (In general the resonance frequency will increase by increasing stiffness, decreasing mass, or both). Very hard to believe though that a fundamental driveshaft resonance problem could make it through the design process and then have to be "fixed" in such a manner. It would be easier to change the ID/OD of the driveshaft to stiffen it appropriately.
However, the driveshaft, as designed, can create other porblems (including driveline vibration harmonics). The as designed configuration does not allow for synchronous running of the ends of the driveshaft (unifrom angular velocity from end to end). By uniform angular velocity, I meant that at any instant in time, the instantaneous angular velocity is the same at both ends of the driveshaft. Just because the driveshaft is "solid" tube, doesn't mean that its instantaneous angular veloicty will be the same from end to end. It is an elastic element that will twist under load (torque). The amount of twist can/will be different end to end and so will the instantaneous angular velocity. The main function of a driveshaft, besides delivering the power, is to maintain uniform angular velocity (in an ideal situation constant velocity) end to end.
To achieve uniform angular velocity through the driveshaft, the ends of the driveshaft must:
1. have the same included angle between the driveshaft and the input (transmission) and the output (differential).
(One of the reasons why setting the pinion angle correctly is so important when installing a rear end.)
2. have the yokes lie in one plane or in-line with each other so as to constantly maintain the same included angle between the drive and driven ends. If the yokes are not in line, it will be impossible to maintain the same included angle between ends of the driveshaft. If the included angle is not equal, the instantaneous angular velocity will aslo be different from one end to the other.
So you might ask, why care about uniform angular velocity between the ends of the driveshaft? Because to have non-uniform angular velocity literally means that at any instant in time the applied torque load will be amplified through increased twisting (angualr deflection) of the driveshaft, increased torsional stresses, forces and vibration - not a good thing.
May never know why, but it sure seems like something we all need to fully understand.