Yes, that is better, I really appreciate it.
You might look at OEIS A271116. It also generates all primes and some pseudoprimes to base 3. I don't know how the timing /efficiency compares to your method, but you can see the Mathematica code there. Ref: https://oeis.org/A271116
I am not looking for anything, just trying to assist Mr Gmirkin up at his higher range. I understand the code completely.