(Old review Title field that was silently removed per typical Mozilla doctrine of "rip away longstanding useful features Because Modernization":) == Meta-search with selectable engines avoids tracking/profiling and infinite Google CAPTCHA loops on Tor! ==
Thank you so much for making this search plugin available for Mozilla-based browsers! I was looking for a solution to the problem of Google often entering (seemingly) infinite CAPTCHA loops when trying to use it from Tor Browser, and searx.me is by far the best solution I found. I love that you can set which search engines to include in your results (although I wish that in addition to being able to set them via "preferences" in a cookie, that there were the option in an Advanced search to temporarily select a set for a particular search, and to allow setting them via an HTTP GET query, while retaining the setting of sending actual query keywords by POST to avoid logging, so that it'd be possible to retain settings across sessions in cookie-forgetting Tor).
Most search engines other than Google do insane stuff like by default, and usually non-defeatably, assuming that you misspelled query terms (e.g. assuming you meant "or" instead of "Tor"), and/or giving you results that merely include words that include your query terms as substrings (e.g. giving you a paper by someone named Cowling as a top result when you're searching for "cow" articles). Unlike other privacy-centric search engines like DuckDuckGo, you can turn off all the engines that give crazily non-relevant results, and end up with a meta-search that doesn't blithely waste your time.
In my use of Searx so far, I've found that Google is the source of most of the results (after turning off the bogus-results-returning engines), but unlike with using Google directly, they will not be tracking and profiling you, even if you have to be logged in to a Google account due to YouTube or other Google services, since Searx does the querying on your behalf. Also, if you're trying to use Google from Tor, you no longer have to deal with infinite image-recognition-based CAPTCHA loops (or just blanket blocking without even the opportunity to try a CAPTCHA) because of "unusual traffic" from your Tor exit node.
One of my only complaints about Searx other than the lack of ability (unless there's some way that's undocumented) to select search engines for a query via HTTP GET parameters is that unlike when using Google directly, you can't set a preference to get 100 results at a time. Also, the "next page" button doesn't give any indication of how many pages of results there are.
In any case, thanks again to the developers of Searx and this search add-on -- y'all are doing a great service to privacy-valuing users worldwide.
P.S. Avoid the "searx.me" search add-on, which was last updated in 2015 and reportedly stopped working in 2016, and stick with this "searx.me 0.9.x" one (the "0.9.x" refers to the version of the Searx codebase, which is open-source and is hosted on various different instances worldwide in addition to searx.me).