Everyone interested by this system, you need to be aware that it’s a potential SEO catastrophy.
A search engine is not going to see any of the translated content, because it does not use scripts.
Not to mention your content run through another third party server, with possible performance impacts.
I tried to fact-check the above as much as I could. Feel free to show me wrong or incomplete.
EDIT NOTE: Read Brandon’s post behind. I previously stated that there was a risk of rating degrading by search engines but as long as you keep using only localize.js, it’s not going to happen.
Sites localization is a hard task. It’s very often under estimated (as badly price-estimated), and proves to be a nightmare to maintain. Even on platforms who are known to bring cool solutions on many things (Wordpress ie) localization is expensive (yearly paid plugin, nights to set up) and complicated. Requires a lot of clicks, difficult and long to maintain, affects the core of your site making deep changes of anything from a bit delicate to impossible.
On big developed CMS like big Drupal, localized pages very often end up to be duplicates. Make a change of structure in a view in English, make the same amends on the localized page. Not how it’s supposed to work, but close to production reality.
At my studio, on top of thinking/making a tool to manage static pages (the goal is to never touch the static pages Webflow exports, but manage them, with domains, urls etc. For the moment, we use a Drupal for that - because we’re transitioning a big Drupal site into Webflow+Drupal before killing the latter, later - it intercepts everything, check the existence of a static page and displays it). We have a topic about bringing localization to this static content.
Here is where we are (I translate what’s below from a developer I work with, syntax can be weird):
"First, we think it should be based on the universal .po format. It’s the translation standard of the unix world. The principle is a source file containing English as a pivot file ( the pot. ) and one additional file by language (eg fr.po es.po po.po ) .
To manage it, tons of tools exist. From the specialized in publishing ( Poedit available on OS X , http://poedit.net/ ) to utilities to reintegrate modifications to already po translated files.
In the long list of these tools , there is a Python-written suite (and therefore running on OS X ) that is fairly complete and especially contains two utilities , one for extracting a jar file from an HTML source, and the other to produce HTML files, taking for input the English HTML file and a po localization file.
The goal is not to do this every time but to integrate it into our static site tool. Basically the principle
would be:
        1. generate the POT of every pages of the site, copied into fr.po, de.po etc
        2. give the PO files to client for translation (and building a tool to do that online would be a good idea)
        3. get the po files back and store them in a /po directory
        4. then look if a translated page exist before displaying any page
        5. display the adequate page"
I know this is going too far for most Webflow users (I don’t master everything I wrote, myself), but it’s the right occasion to share this. A screen cast has been made to show this in action to me, on my webflow pages: http://arnumeral.fr/partages/663bdba97560c3b834ef7976bf16a22d-vincent.ogv