It's funny that people a few posts back are wanting to parse the word "usually" in the Sage Advice Compendium as if it were an unearthed scrap of the original gospels, and now it's "extra-textual support."
Except that's not what happened, at all. Instead, people read 'usually' as the word usually, and treated it that way. You want to delete that word as not saying anything while sticking hard to other words -- in other words, it's not us doing the funny. Either deal with all of the words or don't, but don't try to make that other people's problem. My interpretation, for instance, fully encapsulates your own -- I allow for hiding to achieve surprise as the rules provide. That I also read the other words and see that this isn't the only way to achieve surprise isn't selective parsing.
And, no one here would argue that Sage Advice is anything other than 'extra-textual.' I mean, it definitionally is. Why would this be a bad thing? The Federalist Papers are extra-textual to the US Constitution, for example, but still very important in understanding it. Being extra-textual just means it's not in the text of whatever the focus of discussion is. It's a positional statement, not a value statement.
The rules as they're written are clear : there's a rule printed that specifies how surprise should be determined with Stealth versus passive Perceptions checks, there is no alternative rule given anywhere in the rules, no module published by WOTC that ever sets up a surprise ambush meant to be ran on anything other than on a Stealth basis, and no argument that doesn't try to pluck words or sentences out of context that supports any other meaning.
I mean, there's the big, starting alternative that says "the GM determines who's surprised." Scratch that, it isn't an alternative, it's the top level rule, under which the hidden rules operate without superceding.
It is not the case that all interpretations of the rules are equal, and it is not the case that the Sage Advice Compendium supports any other interpretation of that rule. Here's the intro to that Compendium..
Sage Advice Compendium:
In other words, you can interpret things however you want at your table, but Sage Advice offers the official guidance on how to interpret the rules. It offers no alternative procedure to determine surprise, and it's clear that it presumes that any creature surprising you is hidden:
Sage Advice Compendium:
No one disputes either of these statements you quote.
And it also makes it clear that you're "usually" surprised by failing to notice foes being stealthy, although you can also be surprised by foes with an "especially surprising trait" such as with the gelatinous cube, which is a case of a specific trait over-riding the general rule:
And here you're mixing an matching. The Sage Advice just says "usually." You can also be surprised by a specific rule, yes, no dispute, but the "usually" doesn't limit things to that. Nor does the actual rule, which is "the GM determines who's surprised." There's lots of other rules to give the GM tools to determine that, and it's up to the GM to decide how to use them. The 1/2 a paragraph under the surprise heading that talks to the usual case of hiding doesn't remove the GM's authority to decide, here.
Monster Manual entry for Gelatinous Cube's "Transparent" trait:
Note that the cube doesn't get some advantage to a Deception check - a creature is just automatically surprised. This fits exactly into the modular rule structure envisioned by 5e : the general Stealth-based rule applies unless and only unless some more specific trait indicates otherwise.
No dispute. This doesn't reinforce your argument, though, as it's a specific rule for only that monster, and such specifics are already covered in the general rules as to how they operate. To clarify -- you can be absolutely right about surprise and this works how it does; conversely, I can be absolutely right about surprise and this still works how it does. It doesn't support or detract from either argument. Hence, orthogonal.
Sage Advice Compendium:
You would have to reach so far you'd have to stand on your tippy toes to imagine any of that means anything else. I think some people who struggled with interpreting the text of the rule before the Sage Advice Compendium came out established an early
idée fixe about what the rules meant, and having envisioned a square hole, they're now determined to stuff the round peg into it. But no matter how much you squint, it still doesn't fit.
Oh, and you talk about irony!
Seriously, if we parse that down to something more simple, it reads:
"To be surprised, you must be caught off guard, usually because
condition A occurs or
condition B occurs.... "
Okay, here's the rub. You read this as "it's usually condition A, or it's condition B." Other read this as, "it's usually condition A or B." The punctuation here helps split out the different readings. To you, it's mostly A but sometimes B, but never anything else. To me, that reads it's mostly A or B, but can be something else. Both are fine readings of this specific extra-textual advice (swidt). It's only when you bring it back to the rules that your reading doesn't jive well with me, because the rules say the GM determines and we already have rules for how GM's determine things. The bit about how stealth works really is just a restatement of the general rule of how GM's determine things that deals with hiding. I don't even need to consider gelatinous cube special abilities because they already trump general rules and so don't support or detract from either reading. I just need to look to the general structure of 5e, the tools provided for GMs to determine things, the framework for play, and the fact that the surprise rules start off with saying the GM determines surprise. Viola! So long as the GM thinks a thing might be sufficient for surprise, it's sufficient.
If you intend to extend the core rules by replacing the surprise mechanic by something of your own creation or by some community content, that also fits into the 5e modular system - but, it's no longer running Rules as Written, even though the rules allow for it. If you intentionally make a one-off determination that you allow something as a one-off in play that determines surprise some other way, that's fine : you're running Rules as Written by default, and you've let your players know you may depart from that for one-offs.
I haven't added one word to the rules, and have specifically references which rules I'm using -- references you have yet to address. Please don't accuse me of houseruling when I've done nothing but point at the rules and haven't added a word.
But, none of that changes the rules or how the game intends for them to be used.
You chastised me above for assuming to know the intent of the designers, but here you are doing it yourself. I assume you have some reference for this, because your cites of the rules and Sage Advice are, at best, open to liberal interpretation.