Use "real" testing data from the wild

[Stefan Appelhoff]

For future reference, we can use this dataset (CC0 license):

https://zenodo.org/record/20044

[Chris Holdgraf]

Or whenever the datasets for the EEG and iEEG extensions get released, we can use that data too!

[Stefan Appelhoff]

Given that we do not (currently) plan to implement a READER in pybv, I am closing this.

For reading, mne should be used.