And the fact remains, when the bot does it, it is not.
As noted in the original post, the vast majority of moderation actions are done by the bot. Perhaps most of those are warranted, and perhaps most of the unwarranted ones are only a brief kick and not a suspension, but it's not perfect, and mistakes will be made. The question is whether and how to improve on it.
Reasons have been stated for why making [part of] the bot's configuration public might be an improvement, and reasons have been stated for why making [all/a large part of] the bot's configuration public might be damaging. Seems to me there's quite a bit of room for more transparency there without risking the stated downsides, but that's my opinion.
That's how it should be, but this whole update to the chat moderation system became necessary because that was not always [perceived to be] the case, perhaps especially in recent months. The perception is important, so making visible changes is important.
The fact that the changes to moderator selection, oversight and retention have stopped short of letting [some particular] old moderators go does not mean there wasn't improvement. The fact that there was improvement doesn't mean there's no room for more. Airing the past's dirty laundry in public has been tried, but DE has yet to respond to that, so it seems like a non-starter; we can only try to make sure future infringements of the [newly/slightly clarified] moderator code will be reported and acted on.
Toward that end, I'll echo that the requirement for an opinion to be 'constructive' is just a little too broad to allow a moderator to disrespect a player for holding it. Let's place that particular bar a little bit higher, shall we?
It seems important to note that the chat moderator page prohibits the moderator from doing certain things, such as using 'Any racist, sexist, homophobic, or transphobic terms', which puts the onus on the moderator to make blah-self aware of those terms. That's perfectly reasonable for a community representative. This same onus is on the players themselves in regards to race (where slurs and stereotypes are explicitly actionable), but not to this:
if you're attacking individual players, you're gonna be moderated (duh)
if you're making more broad malicious statements, against 'people that do/think <x>', you're gonna be moderated
if you're knowingly bypassing the filters to try and say things that you know are slurs, you're gonna be moderated
BUT if you just straight up say something that is considered (by whom?) to be a sexual slur, that's NOT on the human-moderation list
That's a pretty big oversight, especially with recent events, to just gloss over. Please don't just leave it at "we said it's non-exhaustive" in regards to this.
I think that 'sexual slurs or pejoratives' should also be on that list if there's gonna be moderation based on them (outside the context of attacks or almost-spellings). If that moderation is handled by a bot, instead of humans, that only solidifies the point. "Modified at our sole discretion" shouldn't mean "configured to enforce additional rules we didn't bother to tell you about." "Nezha is gay" is a useless and annoying thing to say, but doesn't match your current guidelines as actionable. "Traps are gay" is a stupid and hurtful meme, and should absolutely trigger the bot, if not human moderation. "A Gay Guy Plays" is the name of one of your partners, not a slur. "You are gay" is often meant as an insult, but "I am gay" is a neutral statement about oneself. If you're gonna have the word trigger the bot across all contexts, then at least include [potential] sexual slurs in the list of actionable items.
Humans being over-zealous in defining what a slur is, is a problem. The bot being similarly over-zealous, with no feedback and with the only faintest idea of context, is a much bigger one. If you solve the problem of slurs by not addressing it, and secretly changing the bot, you do so at the cost of severe damage to your credibility.