And yet he was a front line character much of the time and managed to live to old age and retirement.
Hold that thought.
Often, but not always. And the cleric didn't always have enough resources to prevent save or dies- imagine the (not uncommon) scenario of poison gas that gets the whole party (of 8ish!).
So we always sort of treated the entire party dynamic as the party exists to keep the cleric alive, and in turn the cleric exists to keep the party alive. Only a supremely powerful character could go anywhere without a cleric along and not expect to have a very short lifespan, and even then that was usually a combination of 'cleric in a bottle' in the form of healing magic items and healing that opens up with higher level paladins, druids, etc. That is because, as you noted, the only good answers to poison and disease were magical, and without clerics you are going to eventually die to one bad situation.
But then you ask me to imagine the "not uncommon" scenario of a save or die gas cloud, which would have then and does now strike me as an extremely bloody minded challenge for a DM to pit against a party. There are in fact no save or die gas clouds in Tomb of Horrors for example (that I recall), and the most bloody minded challenge in the whole dungeon is the sleep gas. If you consider save or die gas clouds "not uncommon" it's not a wonder to me that your impression of the game is one of completely random and unavoidable deaths and so the character you play is basically as good as any other. I think I'd gravitate to only playing Dwarves in a game like that.
Which gets to the idea of how you managed to get a melee fighter to high level without a CON bonus to speak of, and apparently not a lot of clerics. You mentioned early DM fudging, but I'm wondering even more about the mix of challenges involved. You're clearly not trying to face anything like the G series with its loads of giants trying to squish you flat, or anything like DL with its save or die dragon breath weapons. (I'm not necessarily saying everyone played those modules, but remember my assumption was that homebrew games resembled modules or campaigns to some extent.) The real sort of fudging that I think is going on here is more like what I talked about with making a character useful through campaign and encounter design. Then again, this is apparently a game where despite not throwing a lot of difficult combat challenges at the party - few 16HD hydras or yagnodaemons for example - the DMs feel perfectly free to hit the whole party with regular save or die gas traps...
We didn't use henchmen much. With ony pcs and porters, one bad roll was indeed enough.
We had tons of henchmen. In fact, it was possible on a game night where most of the group wasn't showing up, to do ensemble or troupe play where one or two PC's went off on their own and any player that showed up who didn't have a main PC available could play one of the henchmen. Henchmen were essential for providing valuable skills like healing, magic, tanking when key PC's weren't around or were down or out of resources or whatever. The could also guard the baggage/camp if you were going into a dungeon.
Anyway, we're still stuck on my assertion of 'viable', despite my lengthy attempt to explain it and despite the fact that you don't dispute my analysis of the rules.
What I'm not sure of from my end of the conversation is why you think a statement like "There is more to D&D than your ability scores" or "I had fun playing a thief and saw others have fun playing a thief." is a refutation. We agree on both of those concepts, and I don't think either harms my assertion about viability. There is more to D&D than your ability scores, and though you don't define what you think that "more" is, it still remains true that even if your goal is narrative ability scores help that goal by ensuring character survival which in turn ensures continuity of the narrative. There may be more to D&D than ability scores, but the structure of AD&D and to a lesser but still large extent BECMI weighted all the viability of a character to having one or more scores of 16 or higher because it was only at that point that you got advantages in play and those advantages while they seem small were in fact enormous when you start doing the math. Consider that the fighter's most powerful and impactful class ability before weapon specialization showed up and broke the game was the massive advantage they got from 18 STR or 17 or higher CON.
My numbers don't go away just because we both agree you could have fun despite them, nor do they go away just because we both agree that skillful play (by the player) and the attitude of the group could overcome bad design. My point is, even so, "despite" and "bad design".
This is in contrast to say 3e which had advantages start at 12 and linearly increase and had well defined advantages for all ability scores that applied to all classes, so that while pure optimization still might go for Jack One Big Hammer, a broad range of comparatively low scores (12's and 14's) was still plenty viable.
In the context of the "OSR Gripes" what I'm essentially asking is, "Why would you try to have fidelity to older editions exponential and very top loaded ability scores and not utilize the obvious improvements of 3e's ability score bonuses"? I mean, yes I can agree we all had fun in the '80s, but do we not all agree that there were bad design elements? Why are we building games that sell themselves as faithfully recreating the bad design elements?