I've personally fixed ghost shifting from 2>3, 3>4, 5>4 and 4>3. The 5>4 is the most consistant one as the milage increases on a bike and can be the one most predicatable. 1st gear is the only gear that is never involved. The others are more random. "It randomly shifts like a ghost caused it." Who ever named it that, found a good name for it. But that is just symatics.
What makes it elusive to us GW riders is that it happends 5>4, 4>3, 3>2, 2>3, 3>4, 4>5. So there are 6 conditions but really 12 since it can either slowing the rear tire down or speeding the rear tire up. While doing any of that, it can "hang" in a nuetral position between gears. That is why some are able to pull the clutch in before it goes into another gear or possibly back into the same gear. Often that disrups the ghost shift and causes it back into the gear it came from. In either case all those conditions give a completely different feels to the rider, and depending on the road conditions and angle of the bike at the time, can be very scary and dangerous. All of the above conditions occur on acceleration or on the "power" side of the dogs. My yellow '03 was more likely to do it on the 2>3 but also would do it on the 5>4. Yours is unique because it does it on the coast side of the dogs. I would still call that ghost shifting since it os still doing it without the rider causing it. That then makes 24 conditions .... elusive to any rider. This elusiveness is why so many riders blame their transmission problem on "miss shifting." On a GL1800, miss shifting usually occures on the 1>2 shift while passing through neutral never really getting it in gear. What make ghost shifting different then miss shifting is that the transmission is actually shifted into its intended gear. The rider pulled the clutch in, shifted the foot lever up until it didn't go up anymore. In other words, all feed back from the machine indicating it is in gear is correct.
From a mechanics point of view, there is one other thing common to a ghost shifter. All parts are in spec. Yours may not be now from the damage that has occured. For example, your main shaft or counter shaft may have excessive runout from catastrafic failure (sorry, bad speller). But, if measure prior to that failure, and after the shifting problem started, would have been in spec. The other thing common to ghost shifters, if you put Parisner Blue on the gears looking for uneven mess, uneven wear will not be detected either. You will see what looks like bernelling starting on the gear bushings. But it will be in spec. or would have been prior to catasterfy. Put new ones in, with new gears and shaft, then for fun, tear it back down in 500 miles and the exact same wear will be there. You will not be able to tell a bearing with 60k from one with 500 miles. Also, look at shift spindle A. See the excessive wear. Honda does not supply those bearings without a new case, but you and I know how to get around that and install new ones. If you put new ones in with a new spindle and tear it back down in 500 miles, that wear on the new spindle will already be very noticable. See your shift fork shaft, I see wear, measure it and it will be in spec ... same with the wear showing on your main shaft ... it will all be in spec ... all this is common to ghost shifters. Measure the ID of those shift forks, I'll bet those are in spec too.
Another symptom of ghost shifting is Honda's inability to fix it. Anything that I am saying sounding farmilure here? Hasn't yours failed a second time too, or is this all due to rider error? Or in your case, when you put it in gear, are you getting feed back from your transmission that you really are not in gear?
Anyway, how do your gears, shafts, bearings, and forks measure up, any uneven wear patterns? Maybe yours is different then others !!
In the middle of you pix, under everything is what looks like a shaft pointing at 2:30. Is that a broken fork shaft? I cannot tell.