Used the opentts Project with default settings. The voices you will hear in order:

coqui_en_ljspeech

larynx_northern_english_male

larynx_southern_english_female

larynx_scottish_english_male

larynx_glados

I found the gladOS voice funny, hence I included that one, but these are the best sounding to my knowledge so far.

I have seen that with there will be a new mimic release version 3 soon, which has a really great voice. And example with the voice of Alan Pope mycroft_demo. And synesthesiam who's working on these projects said that there will be hopefully a model, which can be trained with actual voice, which would fine tune the voice to sounds more like someone else.