So, basically they have just reverse engineered the already very simple Luhn-10 algorithm, and labelled it as a "new algorithm" to generate tokens which get validated by Luhn-10.

How many lines of code was that? I would be surprised if the guys have written more than 10 lines.