The new type 5 is for characters who are disjointed enough from regular causality for processes operating under regular causality to be completely unable to affect them.
This does not apply to characters who have sufficiently warped causality to where it's difficult to meaningfully affect them. If a character's causality is "reversed", such that all attacks on them were felt in the attacker's past, and are healed in the attacker's future, it would not qualify, as the character with reversed causality is still affected by ordinary causality, even if the effects are changed to be less advantageous for an attacker.
This does not apply to characters who are made difficult to interact with through temporal shenanigans without a causal element being explicitly stated.
This does apply to characters who are on another system of cause and effect (still affected by causality), but for whom being on that different system makes them unable to be interacted with.
Characters who transcend "all systems of cause and effect", and who are made unable to be interacted with because of it, still qualify for type 5.
After consideration then in light of that, I’m gonna say neither that or the arguments above actually address the counterpoints raised
It seems the e=mc^2 argument is being used to set a present but the two situations are not equatable. Creation feats in fiction are inherently ambiguous in their method unless otherwise stated, which is why we treat them as not necessarily equatable to AP, simply because we don’t actually reasonably know if the method itself would be something that would result in such a extrapolation being feasible. But in the case of say, for example, being stated to be unbound from all forms of causality, the method is already known, and the idea of being uninteractible follows naturally. The two situations aren’t really comparable at all nor does the former give precedent to the latter ruling because the issues are inherently different
Additionally, even if we decide to completely handwave the fact that the rulings address two completely different individual, we still wouldn’t be applying the acausality ruling in the same manner as we do creation feats, because with creation feats, the cause (e=mc^2 based physics being used for creation) being assumed to be false, none of the implications that follow (namely AP that is based on e=mc^2, matter and energy manipulation in addition to creation, and the power being affected by powers which manipulate such equations) are all assumed to also not be applicable. But in this case of acausality, only one specific implication (invulnerability and lack of interaction) is removed if the method (being unbound from causality), we also keep some implications like paradox immunity, instead of actually doing what we do with creation feats and ignoring all implications besides the directly shown (in the case of creation feats, that would be having the power of creation, and in the case of statements of being unbound from causality, that would be Resistance to Causality Manipulation)
The long and short of it is, even if we completely ignore all the issues with such a comparison and assume it does set a precedent despite completely different issues, we don’t even actually follow the precedent anyway because we actually give those statements more leeway than creation feats by giving them an Acausality subset rather than just resistance to causality. So regardless, the current system isn’t consistent with either side’s argument
As for the idea of existing within a series in the same fashion as other characters, I would agree personally that that would register as an anti-feat for whatever individual was said to be unbound from causality being uninteractible of course, but the argument that that somehow follows to create blanket rulings like this one is a bit…. I don’t see how it follows?
Ignoring the issue that we don’t judge powers and rulings existence by how many people have them (Transduality and Power Modification come to mind as having few users and applications), but rather by consistent definition, so the number of characters that would apply isn’t super relevant. I’d also point out that there are plenty of characters who don’t interact in the story due to their nature, and plenty of them are only interactible through specific methods, which would be a feat for those methods, not an anti feat for them in any case.
Mind you, the rest of the rulings seem fine, but it seems pretty clear statements of being unbound from cause and effect should include interaction by definition unless otherwise shown. Sure, that may only apply to a few characters since there are a lot of possible anti feats, but our job is to evaluate those things on a case by case basis, not throw out blanket assumptions because “other series have situations where anti feats exist”. In fact, if the list of potential anti feats is so broad, that shows that the necessary strictness for such a ruling would be there anyway simply because literally any character existing in the story in a normal fashion would already be disqualifie. Additional requirements for evidence would be redun anyway and would be more based in assumptions as vs debaters seeing other verses rather than the internal consistency of the story
because it needs to be stated he is impossible to interact with, but creating the whole system of causality does not grant it. Oh, my. Whoever brought that the statement of interaction should be brought otherwise it won't be qualified even tho he is shown to be outside all systems of causality did not understand the basic nature of this type. 
