Metadata improvements on individual data sources

[obo]

Felix (rightfully) asks for more details on the individual data sources, specifically thinking of ASR:

Something for all the datasets here: Mention level of preparation/rehearsal and whether the speaker is native or non-native. Maybe something about audio quality? Close-speaking microphone or noisy conditions?

Please keep that in mind when adding any datasets and improve the already included ones if possible.

[obo]

@Rishu, could you please take over elitr-testset cleanup and improvements in general? Daniel is not with us again for an indefinite period of time... Thanks, O.

[obo]

Lowering priority, but we still need this.