It's a group check and unless there are some wildly different modifiers between the PCs then it looks to me like at least half of the party passed, therefore the party passes, no worries.
I played a shadowdancer in full plate that always used stealth for initiative once. She was practically invisible, but with the implacable CLANK CLANK CLANK precluding a heavy THWACK to start off every combat. it was pretty reliable strat all things considered.
it was ironically our swashbuckler who was the bane of all stealth attempts, as they were just so unlucky with d20 rolls
I hate group checks for this kind of thing. I mainly only use them for perception or knowledge checks (always fun when one person is oblivious). For group tests like stealth or athletics for a chase it's probably better to either build a challenge out of it so other skills can apply and more checks balance the luck factor, or just let one player be skill leader and make the check with appropriate penalties if part of their challenge is managing the clanky loud orc in plate.
I completely disagree with this. The whole point of those types of armor is that it has a drawback for the benefit. What you're proposing essentially nullifies that drawback, giving full armor with no recompense. Moreover, you're suggesting having stuff to balance the roll but that's literally the point of group rolling. To have a group balance each other out and help each other than relying on a single skill leader. It's also a group game. Having a skill leader feels very contradictory to the entire purpose of the game, imho.
I tend to take one or two approaches. First being "Slowest and Loudest" in that the one worst at the test makes the roll, most of the time with help. Possibly also backed by a setup action.
Second is turning it into an extended test, I'll put up a tracker and we'll see what actions the narrative drags up. With this option a failure is only Stealth Over if it has to be, when there is no other reasonable consequence. So clanker can clank.
That's kind of irrelevant though. Whether it's an auto-crit or fail doesn't matter when dude rolled as low as you possibly can. Might not be an autofail at most tables but it might as well be because chances are that this person didn't meet the minimum roll required.
Very much depends on the modifier, though. Like in Baldur's Gate 3 they do crit fails/successes which is what made me think of this. But say my character is a level 20 wizard with an essentially superhuman mastery of Arcana. So a bonus of +12 to arcana and is presented with a rune that needs to be identified:
Under the crit fail/success system, this genius Archmagus with a knowledge of Arcana in the same ballpark as Mystra herself has a 5% chance of not knowing what the fuck that rune does instead of whatever small percentage rolling a minimum of 13 would get you on that particular skill challenege. If this dude rolled the lowest he can roll, it is and should still be treated as pretty damn good.
And it's ultimately up to the DM, of course, but RAW matters too
in 2e, its a pretty big difference. A failed stealth check bumps down your stealth status to "Hidden" (What was that noise? Who's footprints are these? etc.), while a crit fail makes you full on "observed" (Tony Tony Chopper style.)
The Assurance: Stealth feat is the lynchpin in many a clunky fighter's exploration kit, since it will pretty much never critically fail, giving them time to hide somewhere or get help from a sneakier friend.
Sort of since what the DM says ultimately goes, but no - a crit fail means your effort just fails no matter what. Now, it may also mean that your acrobatics check ends in you slipping on a banana peel and breaking your back, but it doesn't have to be dramatic.
So, crit fail means that no matter how skilled you are, you have a 5% chance of failing anything you attempt (without advantage, lucky, etc. anyway)
Plenty of people do, it sucks. Nothing like having a bonus large enough to still pass even on a one but because someone likes more critical failures you fuck up at doing something that should be comically mundane to your supposed expert 5% of the time.
9 times out of 10 this is unfeasible. Armor gives disadvantage because it's cumbersome and makes noise on its own. It also takes time to get out of.
Most situations requiring you to stealth aren't going to have time for you to get out of your armor, pack it up quietly, and then start heading out. Even if they do, you've got to worry about what's up ahead. If there are any enemies then you've now removed your main defense and made it easier for them to kill you.