Why do you think that? (I'm legitimately curious.)
Not parent and I also consider R to be a surprisingly solid general purpose PL but the one reason that I can think of is that formulas are part of the language. Formulas are extremely well-suited for data analysis but odd outside of that.
Good point. I'd add complex numbers and "everything is a vector" to that.
Of course that doesn't make it a DSL. It simply means that R was designed with a particular application in mind. So was Perl (regular expressions as first-class citizens) or Javascript (DOM manipulation). Not to mention PHP.
That is true. Also Erlang with the actor model, and Go with its goroutines for network services.
This is why I said "de facto", not "by design". Yes, you can use it outside stats domain, but it will suck big time for that. PHP used to be like that as well, but later versions look more usable for general scripting.
No R is not a "surprisingly solid general purpose PL". Not in comparison to better option such as Ruby, Python, Perl etc.
But yes in comparison to Stata which is its main competitor in the space where I am using it. This does make it possible to write much more modular, legible, and well-tested R code than Stata do files.