Well sound travels somewhere around 1128 ft/s (cheers wikipedia, I'm used to working in meters!) depending on air temperature, so since the path length difference from source to both mics is around 14 feet, you can work out the relative delay caused by this to be around 13 or 14 milliseconds.
In audio you've got the Haas effect, which states that if the delay between two sounds is less than 30-40ms, the two sounds will be heard as one. If the delay is greater than this, two distinct sounds will be heard (i.e. if it's recordings of the same source, you'll hear an echo). So while the path length difference between the source and your two microphones may cause some filtering effects (comb filtering), the delay shouldn't be enough to hear an echo.
Although I don't have any experience in using the c74 shotgun mic, I can't imagine there would be any latency in the microphone outputs that could be causing your problem (they're all analogue, so their reaction times should be negligible in this regard really).
What methods are you using to record the audio from both mics? It's possible if you're using two different recording devices for each mic that one is introducing a delay, although since you say only some clips exhibit the delay, I imagine this probably isn't the case. Assuming you have the audio recordings in two separate tracks, it might be worth fiddling around with adding 20ms or so of silence in front of one of the tracks in an offending clip to see if that improves things for you - also listen to the audio tracks separately to make sure there isn't a physical echo that's been picked up by a mic (more likely on the SM58 ).
Failing all that, give me a few more details of your setup and location to see if I can think of any other possibilities!  | 
Linear Mode
