That doesn't work if you're encoding with x264, since it doesn't do sound. You can, however, do something like below, open the .avs in VirtualDub and save the resulting WAV. Which still probably is more work than just delaying the audio, but hey, there's more than one way to kill a cat...
Code:
a = avisource("1.avi")
b = imagesource("2.png", end=300, fps=a.framerate()).converttoyv12()
audiodub(b, blankclip(b)) ++ a
Yet another way to do it is by abusing mkvmerge's append function, but then probably need to generate silent audio and then encode it with settings identical to the existing audio stream.
In the end, you probably are best off by just appending the video in Avisynth and delaying the audio in the muxing step. Except if your video is VFR, because then you need to either shift the timecodes, or append the video in mkvmerge. Oh, who am I kidding. Do it whichever way you want, as long as it works. Just don't append AVI segments in Virtualdub, because that doesn't synch the audio. :V