I am not using much Korean here on Lingq but I have actually taken Korean as a class and followed the sounds and read the Hangul and I will say that I absolutely HATE transliteration. It’s not consistent as someone said above, at least to how I would even interpret something properly transliterated.
Case in point, two artists names, Lee Hyun Do and Uhm Jung Hwa. This spelling is common but not correct at all to me, if you want to be phoenetically accurate to how I had learned to associate the sounds it’s Lee Hyeon Do and Eom Jeong Hwa. Those are VERY different so when I suddenly saw how their names were actually spelled I was really pissed as I put in different Hangul for search results.
Japanese is much easier because a lot of those sounds already in English so there are close approximations. This doesn’t work though in Chinese and Korean.
Along these lines Roku needs NON English text support when you are searching for Korean and Chinese movies and actors and actresses.
I already discussed it here, but stemming seems to be at least partly solved in programming, yet we get many duplicated words that should not be duplicated. I understand creating a separate word for different verb endings, but particles always act the same way.
Devs, please do something about this so you can take my money.
Fantastic post - can’t believe I came so late to this thread. I very much agree with #4 and #5 here.
As for #2 though, I think it’s sort of an 어쩔 수 없다 situation 
Naver is by far the best English-Korean dictionary available online, but any bilingual dictionary can’t/won’t contain every word (especially when we’re dealing with rarer homophonic hanja vocabulary, etc…) I think that the current resources on LingQ are probably the best they can do, and more advanced learners who have trouble finding a term in the standard Naver English dictionary should just take up the challenge and start trying to read/understand the native Naver dictionary (or other good resources like 우리말샘, or whatever.)
Resurrecting this thread to say that it would be nice if you could introduce Korean fonts to the web reader. Not a critical improvement, but still it would be a nice addition to have.
Also, a wider range of TTS voices would not go amiss. There are only three currently and the two female voices are quite robotic. The only natural sounding voice at the moment is male.
@TerraEarth
It’s been a while since your post but regarding hiding the progress bar:
Is there any chance you could share with us how you managed to hide it? I wholeheartly agree with you that when reading longer (especially difficult) texts that this is a huge distractor and demotivator.
Thanks in advance.