At first I thought there was a golden number but experience and discussion with other GMs as well as looking at different designs have proven to me that there isn’t. I haven’t seen yet studies about brain performance that could point in an specific direction, though there seems to be a limit to how many things we pay attention to and why. (I’ve been watching the Mind Games show on netflix)
So instead of stopping there I’m thinking now it has more to do with being able to “replace” the tools our mind commonly uses to deal with reality with several procedures and mechanics we use to deal with the fiction, according to the game we’re playing. Most probably the amount of mechanics each of us personally uses has something to do with how much we need to replace to feel comfortable, as it always feels like a game instead of a chore.
So, while there are definitely people whose IQ will influence how many mechanics they can handle, most of us will take hold of as many mechanics as we feel comfortable with, getting them together in procedures/sequences our brains can remember easier or ditching enough to get the game experience we’re familiar with or expect from the game.
While I’m using the term “mechanics”, I’m actually dropping in there together things like rules, techniques, ritual phrases, ways to convey spatial positioning, sequences of procedures that link these together and ways the group exchanges information on everything related to the game.
So, for example, we have ways to:
-Build the fiction. From detailed setting books to monster lists, from having the GM create everything to brainstorming it with the players, or having them create it on the go, etc.
-Decide what and if something constitues a challenge and the outcome of characters meeting them in different ways. Like adjudication, levels of difficulty, dice, etc.
-Define the odds of overcoming a challenge like ways to affect dice probability, bonuses and penalties, basically all the math used.
-Establish and communicate changes of spatial positioning in the fiction. We can go from minis movement rules to words like close and far, to LARPs and more.
-Establish the limits of characters ingerence. From power scale to inventories, attributes, characteristics, skill definitions to random tables of how things can go wrong when using them, etc.
-Reward things that we find good, interesting, fun. From simple laughter and gestures to XP, fanmail, and other in-game resources.
-Keep track of narration and coherence. From communal memory to notes and rules to establish scenes or acts.
-Introduce unexpected, interesting twists into the story, like GM prep, making up new challenges from consequences of previous PC actions, random encounter tables, etc.
-Deal with social situations in the fiction, like roleplaying the characters, rolling dice on social skills, etc.
Help me here, I’m probably forgetting something 
Anyway, my point is that as a designer you can cover as much of these things on your rules, procedures or game advice. The GM/players will fill the rest, but you always have to consider that they will use their own background and bring their own expectations whenever your game doesn’t explicitly offer options to cover for some area, and they even misunderstood,drop and change things in areas you provide solutions for, just to match them to their expectations or background. Meaning that you either have to be explicit and use aids to explain precisely how you cover each area, or go for familiar enough mechanics and procedures to convey the game experience you want the players to enjoy.
Opinions? Arguments against this? Maybe there’s actually a limit to the number of mechanics in use in a game? Would it be better to get close to that number or just design modules the players can use or not or switch around?