As you care enough to keep posting on the topic I will try and be Patient with you. I want to see if I can get you to understand that judging something by the incorrect standards will result in unrealistic expectations. Same motive for me, I want to try to convince you that gameplay issues are caused by gameplay problems.
Your targeting MM because it does not meet your expectations which your making with without consideration for how MM functions. Lets take a stab at it.
Hordes66: No. I judge match quality in the early game, which tells me how well each player knows their champ and their position relative to what is happening on the map. I look at their cs, early health, warding, pings etc. the first thing I need to know is how many apes Riot's shitty matchmaker stuck me with. There is literally no way to judge match quality based on the criteria you described without watching the entire match.
I am still waiting for a reply to
Kai guy: "Ok. Do you know that's not the metric used by MMR match making or even Elo? And why?"
Also dude, it is unrealistic to think Riot can accurately predict Champion selection prior to lobby creation. Contrary to what Hari Seldon and Dr Asimov did in the foundation series, that's not a realistic expectation for the fields of Probability and Statistics. Its a bit irrational to think that mathematics has achieved a level of predetermination rather then estimation. Look up James Maxwell's work on gases or look into Maxwell-Boltzman Distribution a bit. The TLDR is that to accurately measure all the particles in a gas was found to be unrealistic so finding what is "most likely" was adopted instead.
As simplified as I can make this, Randomness is random. Thats why MMR is a Mean average not an absolute value, its to account for human randomness. Thats why Raiting in Elo never locked and would always adapt to new data, humans change over time. Hopefully you can see an issue in treating humans as a constant value variable rather then a variable value variable?