All I have seen of Samarin are academic articles. I have had a look at a few of them in the past few days. I have not read his book that you refer to. Looking at this article from Samarin,
https://tspace.library.utoronto.ca/bitstream/1807/69110/1/The linguisticality of glossolalia.pdf
it would seem unlikely that Samarin was analyzing these samples to determine if they were real languages, but would assume they were not real languages and analyze them for what he thought were characteristics of glossalalia. Samarin even rights that xenoglossia, speaking in real languages one does not know, is not of interests to linguists unless it is a dead language, which would allow linguists to know the pronunciation of ancient languages. His paper makes it sound like his focus was different from what you seem to be arguing for. is the book different?
And for Samarin, 'glossalalia' does not mean what the component parts of the word mean in the text of the New Testament. For Samarin, 'glossa' is not a language like it is in scripture, but a kind of regressive speech. 'Glossalalia' is made from words that show up in Acts 2 and I Corinthians 14, but in the jargon used in this subset of sociolinguistics.
For me the more interesting experiment would be to take samples of speaking in tongues and samples of naturally spoken not-well-known human languages and have linguists analyze both and determine which ones are just examples of natural speech. I think taking that sound at least superficially like human languages to include in the samples of speaking in tongues would be appropriate. Also, the natural language samples could be taken from people praying from the same faith-communities the 'tongues' samples were taken from. This might produce two samples with similarities of tone and style.