DeLux, I question the premise that "hitboxes are how the score is altered in the game we play". That assertion and the following logic implies that the goal of the game is to force the opponent's model into hitboxes, which is not strictly accurate. The goal of the game is to remove the opponent's stocks, and forcing hitboxes onto the opponent is merely a means towards that end. When a hazard deals damage to a player, it does indeed alter the scoreboard because it is universally accepted that a player with more damage is generally closer to losing his stock. However, this is NOT the only way in which a stage can alter the scoreboard.
Given that the win/loss criteria by definition has a percent factor, percent does alter the score. Obviously the other way a hazard can affect is by removing stocks, which would be intrusive in the same manner. So the distinction you're trying to make with that isn't meaningful since it applies anyways.
It is true that we can make a clear distinction between the influence of the Smashville platform and the influence of Halberd's hazards. I argue that we must also make this distinction between the influence of the Smashville platform and the influence of Yoshi's Island's Support Ghosts. They can save stocks that are otherwise sure to be lost. They can place players in bad positions they would not otherwise be in (lifting a Falco up-b, or lifting a Marth just as he uses up-b). They can even take stocks of certain characters. If we do not limit our evaluation to influences involving hitboxes, it is clear that this is a form of player vs. computer interaction. When a stock is saved, the stage is "interacting with a players hurtbox which happens independent of an opponent's hitbox and [...] alters the score in the win/loss criteria". The Support Ghosts "exist independently as their own entity". All of the things you said about stage hitboxes apply to the Support Ghosts as well, except that their influence is not exerted through a hitbox. Nonetheless, that influence cannot be denied. And whether they will appear is random on top of all that. The stage is certainly no more fit than Halberd for determining the more skilled player.
If we look at it as the definition of the premise I'd ask you this question:
Does the support ghost actually alter the score?
Let's look at this literally:
What is the person's stock/percent count prior to touching the support ghost? Hypothetically, it's X total
What is the person's stock/percent count prior after touching the support ghost? Hypothetically, it's still the exact same total
So from a literal standpoint there's an argument to be made that the ghost literally does not alter the score.
From a hypothetical "well he SHOULD have died" a couple things:
Saying one SHOULD have done anything again leads to arbitrary discussions about what should/shouldn't have happened. What if the opponent hit them while recovering to give them their jump back etc.
Second: saying the support ghost always functions in that manner would not necessarily be correct. For example, when the game is going to time, there is a specific period of time where the support ghosts function would be negligible in the context of the game going to time since the opponent couldn't get to the blast zone in an allotted time anyways.
So realistically it's literally justifiable as such, but I would probably agree with you that it defeats the spirit of the rulesset. Alternatively, I'd just concede we could just ban the stage and insert PS2 in its place and remove YIB and Frigate Orpheon (since Frigate also operates on similar questionable grounds) for a functional 9 stage list. More people like YIB rather than PS2 so I don't mind it personally. You could even just axe YIB for that reason and then axe another stage to operate on a functional 7 stages, etc. But that decision is up to you guys as a scene. I just thought I'd offer you both lines of thought on the matter since it's not the first time I've seen it. It's a gray area either way but it can be accounted for or discounted depending on how literal you want to be
