It is still useful to say "prime numbers don't end in 0, 2, 4, 5, or 8" just as it is useful to say "consecutive prime numbers in any base are less likely to be followed by a number with the same least-significant digit". There are special cases at the bottom for both statements.

Likewise in base ten, 2 and 5 are prime. But this is a statistical argument. So that's cute but not significant?

"100% of the base-2 primes ending in '0' are followed by a prime ending in '1'."