It's funny, the spec they references[0] makes no mention of parsers or nesting. But RFC 8259[1] clearly does. Moreover, I notice they were both published in December 2017. Why two specs for the same standard?
As is almost always the case when there are multiple specs purporting to describe the same thing: politics.
The Ecma spec is meant to be a spec defining a format and its semantics; the RFC places requirements on implementations of that format as well as describing it's semantics.
There was a rumor that was IETF was going to specify a bunch of new breaking changes in RFC 7195 (March 2014), and because of that (irrational) fear ECMA published the 1st edition of 404 very quickly in October of 2013.
RFC 8259 and the 2nd edition of ECMA-404 contain the same change, a single sentence, specifying that JSON must use UTF-8.
[0] http://www.ecma-international.org/publications/files/ECMA-ST...
[1] https://tools.ietf.org/html/rfc8259