If this is the remaining objection, then I’m not too concerned. 
Interesting take, considering that the active participants in each thread has been different each time. It seems instead that this thread is less of dogpile and shouting match, and more of a skeptical and pensive “hmm, that’s interesting…”. I think that’s a step in the right direction.
Not true. it and them are fully-local arguments to the chain, and they do not capture from their environment. In essence, they are locally defined keywords.
Locally-defined keywords aren’t unfamiliar if you have ever used the as keyword, or the abstract keyword, or the type keyword, which are keywords when used in a certain context, but outside of that context can be assigned other values.
Indeed that has been one of my concerns too. This is, in part, why I’ve chosen to adopt concepts and keywords from natural language—so that if a concept is foreign to mathematics or other programming languages, at least it is familiar to anyone who speaks English.
Edit: The concepts are also familiar to people who don’t speak English, but the keywords are different. German has “es” and “sie,” Spanish has “lo/la” and “los/las,” and I know I shouldn’t trust Wikipedia, but it says Chinese has “tā” and “tāmen.” The universality of singular and plural pronouns makes this an attractive concept to piggyback on. I have chosen English pronouns because Julia uses English keywords.