I agree with your assessment that problems with interpretation and lack of independent review can definitely make it problematic for editors to cite these interviews directly, and we'll have to see whether it is in any way feasible under any circumstances, and if so, what guidelines can be set up.
*What I do think is incredibly important though is that this material has huge value in itself - and every effort to encourage more of the same should be taken! *
In fact we should get as much material such as this as possible, host it, translate it, make it accessible - and encourage secondary academic sources to make use of it. This could work both as a "hack" to get around the issues of citing oral material directly as well as contributing to the effort to expand knowledge of these areas of study.
A very useful suggestion! That should address the concerns quite well, as well as improving the contacts between Wikipedia and Academia.
Kind regards,