If you want to parse more than 4GB of JSON in a browser, ArrayBuffer versus string isn't the most important criteria - you shouldn't parse a 4GB ArrayBuffer as a single chunk either. You can always convert between an ArrayBuffer and a string, but you'll want to compare parsers according to the ability to stream, as well as performance and memory efficiency.
You could look into streaming parsers such as Oboe.js, which specifically support the use case of parsing JSON tree's larger than the available RAM[0]. Then again, when you're loading such huge JSON files into a 32-bit instance of Chrome, it is likely you should look for a totally different solution to your problem.
JSON.parse() only accepts strings.
The library that the article recommends also uses XMLHttpRequest with strings. [2]
The reason I'm asking is the maximum string length in 32-bit Chrome.
[1]: https://developer.mozilla.org/en-US/docs/Web/API/XMLHttpRequ...
[2]: https://github.com/eBay/jsonpipe/blob/master/lib/net/xhr.js