10+ shinies are not a big enough sample size to conclude that.
Also, I've not checked RNG protocol for the SV non-static wild encounters (I've checked for other Gen 8 and Gen 9 generation methods), but if the method used for wilds generates a 'fake trainer ID', the only way to get a non 0/1 xor would be for the save's TSV to match the fake trainer's TSV (in this case, it's at least as rare as double odds, if not rarer)
I hope you can accept my somewhat objective answer.