I have found all of the planned Duolingo languages using a simple trick
I will be using Afrikaans as an example. As you can see in the following URL, "af" represents Afrikaans.
The full list of two-letter language codes can be found here: https://en.wikipedia.org/wiki/List_of_ISO_639-1_codes under the 639-1 column.
These are the ones that worked when I typed them in:
Afrikaans Albanian Armenian Azeri Bengali Bulgarian Catalan Croatian Czech Estonian Finnish Galician Georgian Gujarati Hebrew Icelandic Kannada Kazakh Kyrgyz Lithuanian Latvian Macedonian Malay Marathi Mongolian Punjabi Persian/Farsi Serbian Slovak Slovenian Swahili Tamil Telugu Urdu Uzbek
I believe I have everything, but feel free to tell me if I missed something.
Edit: there is also Sindarin, Dutch (Belgium), Klingon, and Chinese (Traditional)! Thanks, davidvdb
Edit: I also found Dothraki by looking in the source code for the Incubator.
Edit: Here is the source code: duo.language_names = ['Greek', 'Esperanto', 'English', 'Swedish (Sweden)', 'Lolcat', 'Afrikaans', 'Swahili', 'Catalan', 'Chinese (Cantonese)', 'Gujarati', 'Swedish', 'Zombie', 'Czech', 'Arabic', 'Ukrainian', 'Irish', 'Basque', 'Estonian', 'Chinese (Traditional)', 'Azeri', 'English (Pirate)', 'Spanish', 'Chinese', 'Russian', 'Galician', 'Norwegian (Nynorsk)', 'Norwegian', 'Norwegian (Bokm\xc3\xa5l)', 'Turkish', 'Indonesian', 'Latvian', 'Lithuanian', 'Punjabi', 'Thai', 'Vietnamese', 'Italian', 'Hebrew', 'Icelandic', 'Polish', 'Tamil', 'Dutch (Netherlands)', 'Belarusian', 'French', 'Bulgarian', 'Dothraki', 'Telugu', 'Slovenian', 'Croatian', 'Bengali', 'German', 'Dutch (Belgium)', 'Danish', 'Farsi', 'Uzbek', 'Hindi', 'Finnish', 'Armenian', 'Hungarian', 'Urdu', 'Japanese', 'Faroese', 'Georgian', 'Romanian', 'Portuguese', 'Malay (Brunei)', 'Kazakh', 'Albanian', 'Mongolian', 'Korean', 'Kannada', 'Macedonian', 'Klingon', 'Slovak', 'Malay (Malaysia)', 'Tatar', 'Sindarin', 'Malay', 'Marathi', 'Kyrgyz', 'Swedish (Finland)'];
And to expand a bit on what I pointed out in that thread, most of these are just automatically generated based on the language code combinations that they've accounted for in their code / routing rules for their framework. They're basically place holders, and are not necessarily indicative of any immediate plans to begin work on those courses.
Also, it's worth noting since that post over 4 months ago, Duolingo has added 2 "source languages" (as I described them in that post) and 1 additional "target language" for each. This brings the number of combinations to 1760 from the 1580 mentioned previously. I should also mention that the target languages list also contains the source language as well as some jokes (zombie, lolcat and pirate), so in reality there are only 76 targets per source so only 1672 of those are legitimately possible future courses.
No Maltese. I doubt this means anything, I mean look at this: http://incubator.duolingo.com/courses/en/en/status .
English for English speakers. You linked it in the first post on this comment chain. http://incubator.duolingo.com/courses/en/en/status
Try sjn, tlh, zh-TW, and nl-BE. :) And this one will be very useful, if the elves will ever get out of Middle Earth. :)
Instead of giving you a 404 error - they are giving you a nice message. It doesn't mean anything, other than that there is a defined list of languages. They display the message for any combination of these languages. Meaningless. It is probably exactly the same list they use of the volunteering page.