This helped crystalize some of my thoughts on this. You're essentially making a case for GMing as a profession, with the same slightly murky blend of professional standards/responsibilities we expected from certain experts, like doctors or lawyers. There is a set of ethics and a wide variety of available tools, but fundamentally the role rests in a person and not a system, such that an individual must inhabit the function and make decisions that best serve the table.
Since most of us don't get paid, I tend to not look at things in terms of professional ethics but I admit that they generally apply. I do look at being a GM very much as an artistic endeavor that combines a lot of different skills and which people who undertake it should be undertaking with the idea of getting good at it the same way a hobbyist of almost any other sort - basketball player, carpenter, guitar player, video gamer - is striving to improve their craft for the pleasure of doing that.
In terms of ethics, my axiom is "Be the GM that you would want to have if you were a player", or if you like, "Be the GM your players would want to have."
I think that's the distinction I've increasingly found...uncomfortable in other forms of GM advice, or even some games that have very specific procedures for GM side play, or discussions about what a GM is doing...
The last 15 years or so we've seen - following recognition that the processes of play are at least as important as the rules and that the GM role is therefore at least as important as the rules - attempts to formally encode the processes of play with the idea that if you could somehow encode the GM's role into the rules sufficiently, you couldn't actually have a bad game that followed the rules and that all bad play would in some sense result from breaking the rules or failing to understand the rules.
This is about as naive in my opinion as the fetishization of realism we saw in the 80's through the early 90's where so many designers acted like all table problems were the result of rules that weren't realistic enough, and that if you just had realistic enough rules then the game would just work. For one thing, rigidly encoding the processes of play to produce that one game means that the game probably isn't going to have the flexibility it needs for long form stories unless the GM is consciously or unconsciously ignoring processes of play.
Nonetheless, this fad hasn't quite run its course yet so we are still seeing people argue as if stated intentions of the rules are encoded into the rules then by golly that's what the game does, unlike those archaic other games that don't code their stated intentions into the rules and so don't do that.
But what I've increasingly found uncomfortable and is what prompted this discussion is that as the designers start to realize that the encoding the processes of play is neither as straightforward as claimed nor fixing as much is as claimed, the designers are increasingly turning to patching everything with illusionism. I've seen this idea coming from everything from old school CoC, to Indy gaming, to OSR. Everywhere I go these days everyone is preaching Illusionism as the solution to every question about how to make a game work. It is to fiction what "Rulings over Rules" or "No rule is a bad rule because the GM can just change them" is to mechanics. System not working? Just apply more Illusionism! Players aren't having fun? Just apply more Illusionism! Adventure is written badly? Apply more illusionism. Patch everything on the fly. And while on one level this does fit into my idea of being a skillful GM in that a skillful GM ought to not be hidebound and should have a big toolbox, I'm really uncomfortable with offering up loads of Illusionism and Fudging as the solution to every problem or as an excuse for poorly written content.
Illusionism run rampant is obviously bad for the integrity of the game, but deployed sparingly, with regard for the temptation it offers, and with an eye toward mitigating its effects on both your players and play, is both a tool and a necessary component of the GM's craft.
If you read my essay on railroading, I don't actually say that the techniques are bad techniques. I say rather that they have to be employed sparingly or artfully, with do consideration of the fact that you are sacrificing player agency and you better have a really good reason to do that and further that if you do it too much, then you go from being someone who is using railroading techniques to someone who is running a railroad.
This is I hope a bit more of a quantitative and artistic approach to the topic than the usually binary good/bad qualitative discussion you normally see.
