Though I think most of VanillaSwirls comments are generally confused, there are some biases that show up in this space. In particular, people rate the intelligibility/clarity of Caucasian speakers higher than that of East Asians even when the audio is the same.[0]

[0]: http://www.academia.edu/6126025/Expectations_on_Speech_Evalu...

Key quote from page 444: In a typical study in this paradigm, participants listened to 4 minutes of a tape-recorded lecture produced by a native speaker of standard American English. Some participants were lad[sic] to believe that they were listening to a North American NS[native speaker] instructor, whereas others were lead to believe that the instructor was an international NNS[non-native speaker].

[...]

[I]ndeed, listeners who were exposed to the Chinese/NNS guise perceived more of a foreign accent and scored lower on a recall test than those who were exposed to the Caucasian/NS guise even though the audiotape they heard was exactly identical (standard American English).

Yeah, that's definitely true. My brother is an applied linguistics grad student, so I've heard this before.

I've had instructors who had accents but could still speak English very well, and many people just dismissed them outright as unintelligible without even trying.

But I've also had instructors who just weren't fluent in English (mostly grad student instructors). When VanillaSwirls said they didn't speak English well, that's what I was thinking of.