I think you have another couple options (even if the first doesn't make sense), including a variation of one you've mentioned:
<1> Leave it as is, flaws and all. Sounds like the silence and artifiacts are substantial enough that this probably doesn't make sense.
<2> Convert the recording to dual-mono (delete the bad channel and double the good one)
<3> Keep the recording stereo for the first 40min, and then cross-fade using...
<a> the good channel to patch the silence and noise for the remainder, or
<b> dual-mono for the remainder, rather than taking the time to patch up the flaws
It's really a personal choice. Do what sounds best to your ears and matches the effort you're willing to put into the fix.
Personally, even with minimal stereo separation, I prefer stereo recordings. So I would do <3a> or <3b>. Which one would depend on how long the recording continues after the 40min mark, how much stereo recording (without flaw) remains after the 40min mark, how many flaws exist, and how much effort I would have to put in for fix <3a>. If I felt it was too much work and/or not worth the hassle, I'd do <3b>. FWIW, I generally prefer a long-ish crossfade (5-30sec, depending) when making these types of transitions -- whether patching or simply converting to dual-mono -- because I find it easier on my ears.
As for avoiding the problem in the future: systematically test your gear to identify the root cause of the problem. First, try to duplicate the problem. If you can't duplicate the problem, things become more difficult. But if you can duplicate the problem, try to isolate the root cause. If you suspect the cable, test with the cable. Then test without the cable. And perhaps test with a known-good replacement cable. If the problem only occurs with the suspect cable in the recording chain, then it's the culprit. If the problem persists, follow the same process with the next most likely suspect component. Etc.