The key is the Source Recordings. If they are the same, then you have a reason for the merge.
My short reply is - option 2. Which in your example leaves the classic “messy half linked” puzzle. I’d merge tracks 3 and 6 as assume that is crowd noise, but the others would need listening to as too big a difference to guess. Maybe end up being linked as “edit of” or just left pointing at the same Works as the common link.
My longer waffle of experience follows… 
I’d merge those of a similar duration, as long as they have shared musical content and recording source. No matter how many years apart the tapes have been picked up.
I get this with concert bootlegs, and this is not much different. I’ve spent a lot of time sorting out Peel Sessions and Pink Floyd concerts.
My rule set I use from the way guidelines have been explained to me:
What stops a merge is:
- Large length difference (15 seconds in a 3 mins track)
- Any edits in the audio
- Any extra speaking from band, show or radio presenter
- Different edits (i.e. the chat from the band being moved from end of one track to start of the next)
- Different sources (No merging a mixing desk with a radio recording)
What doesn’t stop a merge:
- Crackles and dropouts in the audio
- Cleaned up audio (remastering) (this includes “going back to original tapes”)
- Applause less then 10 seconds (treated as as “early fade out” as nothing musically lost)
- Fade-in or fade out
When you have a track that has been split in two, then you have a “compilation of”.
Those intros would lead to having “edit of” versions. Bit like when I get John Peel talking at the front of a Peel Session. I have to keep those separate from the recordings where his voice has been edited out.
Importantly I’d look for the two concerts to have generally the same length tracks as much as possible.
If one person has taken the full tape and chopped up the sections between tracks in a very different way, then it stays separate recordings. But if two different people follow the same general rule of splitting tracks after crowd noise and before a spoken intro - then merge.
If looking at a Mass Merge Script and the tracks all have very different edits, then I don’t merge.
Generally I am working on bootlegs that are known bootlegs of bootlegs. So it can be easier. I would not merge official recordings into a bootleg (Apart from some of those EMI cash-in boxes from EMI who have the cheek to sell bootlegs back to the fans)
Hope some of that waffle \ experience is of use. Many AEs and other Editors have helped refine those above notes to build that guidance I follow. 
