I have been importing Cantonese radio content and using the transcription features.
I have also tried out the quiz mode. This ends with an arrange the sentence fragment and speaking test, both of which are very useful exercises. I have found a challenge with the speaking test, though, as Lingq does not recognize alternative characters as the same.
The transcription tends to use simplified characters and the speaking transcribes into traditional characters. Even with the same system, there are alternate characters for some words. The test marks it as wrong unless the characters are identical. This is a bit like having it marked wrong for using a different font 
or using an American English spelling instead of British English.
It’s actually a more general problem with Chinese as it sees the same word in alternate characters as a different Lingq to learn.
For example the text had 内地 and the spoken transcription gave 內地, which did not match. Similarly with. 能夠 and 能够.
It would be great if you could improve this as this feature is very good practice.
Thanks - Anthony