I am also excited that you offered to help on the project 
Besides the fact that many people do not use the built-in translation facilities all that often, there’s other aspects where human-translated resources work best. First of there’s the accuracy of the translation where people still do better than machine translation. But more importantly the mere fact that the resource does not exist in a certain language means that people will never be exposed to it in the first place. Having the translated resource means it can be exposed via SEO, appears in more ‘filter bubbles’ when doing a search, and it can be promoted by people in their own social media targeting their language audience.
What many people do not realize is how much language separates groups. For instance there is an entire ‘parallel online universe’ when considering the Spanish speaking parts of the world.
Furthermore many resources require a form of ‘localization’ as the links on the resource point to yet more English-language articles, and for these - in many cases - appropriate alternatives need to be found. A good example is Social Cooling, which has proven itself a very popular and informative website.
(Note that Social Cooling by @Tijmen - just his like Privacy Label - is an example where we’ll more likely just donate the translation to the creator for inclusion on their website, instead of adding to our own site - unless explicit permission is given to do so. In return we might ask for proper attribution. These resources may still be referenced on the site, and part of promotion campaigns)
Re: Search. Yes, I think a search facility is needed as the number of resources will eventually grow, making things harder to find. A good categorization and tagging scheme (both native features of Jekyll) come a long way, but if your tag filter still yields, say, 5 pages people intend not to navigate all of them. Besides tagging there can also be a ranking based on popularity and a ‘Featured’ articles, both to highlight the gems in the collection.