I was not at all impressed with the analysis in that thread. It makes a bunch of assumptions that don't feel very thorough to me, but announces them as if they are unimpeachable facts.
Biggest example is this one:
"9. Making up words or misspelling hurts - Words that are identified as “unknown language” are given 0.01, which is a huge penalty."
The code in the screenshot for that looks like this:
// Boost (demotion) if the tweet language is not one of user's
// understandable languages, nor interface language.
optional double unknownLanguageBoost = 0.01
That doesn't match the description of "Making up words or misspelling hurts" at all!
Biggest example is this one:
"9. Making up words or misspelling hurts - Words that are identified as “unknown language” are given 0.01, which is a huge penalty."
The code in the screenshot for that looks like this:
That doesn't match the description of "Making up words or misspelling hurts" at all!