Generally though, is it reasonable for features such as these to be included at the language level? I primarily use ES5 JavaScript so I'm used to pulling in a decent amount of modules (in this case handlebars, underscore, etc.) for things like templating.

But I will be relieved to have a language-standard solution to what feels like such a common problem for the environments in which JS is used primarily, despite the tradeoff that's made in accessibility due to feature bloat.

Unsafe string-munging is too easy and seductive. Doing things the safe/secure way needs to be at least as easy.

And it makes for a growable language, the way Lisp macros do. E.g. https://github.com/erights/quasiParserGenerator This can reduce the demand for feature bloat in future language standards.

I'm excited for not having external libraries to handle string interpolation.

I've recently started a new project in React/Flux, and I've cut away 3 or 4 dependencies compared to previous projects by fully embracing ES2015/2016, as Babel automatically provides the necessary polyfills and compilation.

As for template strings, I think they make sense. Most programming languages have them in someway or another, like Ruby "hello #{world}" or Python 'Hello {world}'.format(world='World').

[1] https://www.reindex.io/blog/you-might-not-need-underscore/

Admittedly you could trivially extend the string prototype to have something more akin to string.format or printf, which was something you couldn't do in other languages for a long time, making it less of a problem.

In python you can save a string like: template = '%s is %s' and _then_ evaluate its output: print template % ('john', 'tall')

    var multi_line = 'You already \
        can do this in ES5.';