Rebuttal:
1) I don't believe in System Agnostic adventures being any good unless they are constructed to be solved almost completely with broad ingenuity or social interaction. They fail tactically and strategically because they fail to take into account the effect of the rules on PC decision making and encounter balance, the viability of different tactical options, between-system differences in relative power differentials between PCs and their opponents etc. etc. etc. In short, System Agnostic Adventures are not adventures, they are at best Outlines of adventures that leave the hardest part up to the actual GM. I'm sure there are exceptions, I speak of this as a general rule.
Maybe nobody has been able to successfully write one yet? I envision it thusly:
The adventure presents an obstacle (combat, roleplay, puzzle, trap... literally any obstacle). The players determine how to overcome obstacle, based on what their characters can do.
No matter what edition you are playing, most options that a character can do are universal - they can always sneak, they can always speak, they can always fight, they can always run away, they can always climb walls, they can always hide, etc.
Yes, in a sense, the version of the game you're playing determines how likely a party is to do one action over the other - if sneaking is difficult to accomplish in Game X but easy in Game Y, the players of Game Y are more likely to sneak by the obstacle than the players of Game X would be.
Here's the thing though - the author of a system agnostic adventure (if he's doing the job right) isn't supposed to allude to the "right way" for the party to overcome the obstacle. That's called railroading, and it's always been a hallmark of bad design. So if the adventure designer has made the obstacle solvable only by sneaking, then yes, they have failed to take into account the effects of the rules. Bad designer. BUT, if problems can be solved in more than one way, then the rule system used to render a success or failure verdict (and that's literally what a rule system is supposed to do) is inconsequential. The players choose an action, the rules are used to determine success or failure, the obstacle is passed or not.
System Agnostic adventures, done right, are only supposed to lay out a series interesting obstacles and developments, shrouded in mystery and all-important "evocativeness". I fail to see what's wrong with such a set up.
Let's take a recommended example - say Deep Carbon Observatory - an adventure that I've used, that I like, and that is critically acclaimed... if you strip out the stat blocks for the encounters (easy task), it's literally system agnostic at that point. There's no DC listed to overcome poisons. There's no "to hit" listed next to booby-traps. An OSRIC player might use a different solution to get past the crawling giant than a D&D5e player would, but it detracts nothing from the adventure (the adventure basically says the giant is invincible anyways). Is DCO a garbage adventure for it? No, it's better for it. It's versatile. I can run DCO with D&D5e as easily as I could run it with a game of Traveler.
You say that the presence or lack of certain rules changes the viability of tactics and player decision making, but I say "so what? Isn't versatility a good thing?"
2) I Played 5e for...10? Sessions until I gave it up. GM quality has something to do with it but a lot of the problems I had with it were baked into the system. I like the relative dearth of abilities and magic in older games because it forces players to be creative or to think outside the box, to set ambushes and to use trickery and genuine cunning. Conversely, 5e loads you up with at will powers right out of the gate which are mechanically more formidable then a simple ambush or cunning ruse, meaning PCs are far more likely to think in terms of winning combat through proper use of abilities then circumventing combat through trickery or ambush. There are nine bazillion other impactful mechanical differences, and stating that one can simply alter the rules to alter the game is an admission that rules are impactful on games and are thus significant. I consider 5e to be a vastly more playable though somewhat dumbed down version of 3.5.
I think your problem is that you didn't give the game a fair shot, maybe you had a bad experience, and that your opinion is heavily colored by other editions. I have run dozens and dozens of 5e sessions where my players were able to use cunning ruses and ambushes to achieve their goals, without resorting to brute-forcing a mechanical solution - you saying that you didn't get a chance to do it doesn't set any of that in stone, and your blanket assumption of all players playing/running the game in the same way you did is a bit narrow-minded.
No offense, but if your argument is that there's no room for creative solutions or imaginative scenarios in non-old-school games, then your group is likely unimaginative and uncreative, not the system. You know what more character abilities and magic gives you? More options and more choice! You don't get some "pick fireball every time to solve the problem" button; there is still much problem solving required (unless you've got a shit DM or a badly-written adventure).
I'm not saying there aren't certain rulesets out there that would be better suited for certain types of players; I am saying that writing off a whole edition as terrible because it doesn't suit your one specific group's playstyle is ridiculously shortsighted and totally unfair.
Amen brother. I tried a stab at converting RPR to Zweihander, the mechanical difference alone is staggering so what constitutes a good flow of encounters is utterly changed, the game has to be overhauled to accomodate for elements like Peril and skill challenges and the difference in ability between enemies and PCs is narrower. If you factor in elements like healing and how hit points work the position that rulesets are interchangeable becomes untenable.
My process for conversion in the past goes like this: strip away existing mechanics embedded in the obstacle to make it neutral (instead of having "DC 15 to avoid waking the dragon", you say "waking the dragon is likely if the party does not take efforts to conceal their approach") > insert the appropriate mechanics balanced to match your system ("25% to wake the dragon if any noise is made"). Voila, converted.
This works for pretty much everything.
Ghouls much stronger in 1e than 5e? Strip the stats. Describe the creature like a ghoul but use the HP/AC/etc. for a 1e goblin. Maybe don't call it a ghoul - that reference is tied to a system already; call it "Shuffling Death" or something. Converted.
Scything blades do 3d6 damage - dangerous to OSRIC but less so for 5e. "Scything blades fall from ceiling - DM discretion for damage"... it's agnostic now. Sub in whatever damage is fair for your system. Converted.
Fireball does 6d6 damage... too overpowered for your characters at their level? Change the damage. 2d6. Converted.
20' wide pit trap? Non-system specific already. Many ways to bypass independent of rules. No need to even convert anything.
Climbing wall in a system with no climbing? Well, that's a system problem, not an adventure problem.
Maybe I'm failing to see some unseen hardship for you guys in all this; an example of something you can't easily convert would go a long way to clear this up, because right now I can't grasp what the fuss is about.
