I argue the answer is no, because there's at least a high likelihood that consent wasn't freely given.
So if your intent was to understand the culture and tradition of these peoples, or their artifacts, it makes a lot of sense to respect their traditions and cultural practices. As you know though things never quite happened this way -- there was a genocide -- extermination of their culture and people. So given the complete lack of respect or willingness to follow this principle historically, along with a genocide, researchers are especially careful to respect the wishes of the peoples whose stories, artifacts, traditions they are dealing with.
To your knowledge, did the people being recorded give informed consent to having those recordings heard by more than a few hundred people doing anthropological research?
Congratulations on having an opinion.
Even by Mickey Mouse standards of copyright, which don't even extend to works this old, these phonographs have been in the public domain for 15 or more years.
Not that this is the point I intended to argue.
> The special context for this is the previous mistreatment of indigenous people and their artefacts; therefore researchers these days tend to go to great lengths to include indigenous people and seek their approval, as a courtesy.
As for the previous (and ongoing!) mistreatment of indigenous people and their artifacts; I don't think that has anything at all to do with the ethics or rights regarding the publication of these recordings. To my knowledge, these recordings were never intended to be private to the band in the first place.
If people who lived near or with your great grandfather killed my great grandfather's people or ran them out of town, and my grandfather had on a separate occasion recorded a phonograph, knowing that it would be available to the recording party and all of their associates, in a language you're studying in present day, the crimes against my great grandfather's people would have absolutely nothing to do with whether or not you should publish a recording he made available in the same timeframe (or in this case, after the brunt of the damage had already been done).