If you're not, then make your own template with DATUM, or join another pool. The censorship resistance isn't established by pools or individual template composers being forced towards being inclusive, it's by the fact that mining isn't gatekept, that blocks are permissive and that there is a massive economical disincentive to reject a valid block.
Also, not mining a tx in mempool isn't censorship per se, it's policy. And although for a longer time, from a LN perspective uniformity in block templates was expected, it isn't sustainable and I feel like we've largely moved away from that delusion since full RBF.
True censorship would more look like this:
I post transaction A at height N
FreedomPool mines transaction A into block N+1.
CensorPool censors transaction A by invalidating block N+1 locally and instead mining block N+1b that doesn't contain transaction A, upon N
CensorPool also mines block N+2 upon block N+1b so everyone will pick that up as the longest chain
FreedomPool mines transaction A into block N+3, upon block N+2
CensorPool again censors transaction A by invalidating block N+3 and instead mining block N+3b, again without transaction A, upon N+2, and another block
DATUM, or join another pool. The censorship resistance isn't established by pools or individual template composers being forced towards being inclusive, it's by the fact that mining isn't gatekept, that blocks are permissive and that there is a massive economical disincentive to reject a valid block.Aat heightNFreedomPoolmines transactionAinto blockN+1.CensorPoolcensors transactionAby invalidating blockN+1locally and instead mining blockN+1bthat doesn't contain transactionA, uponNCensorPoolalso mines blockN+2upon blockN+1bso everyone will pick that up as the longest chainFreedomPoolmines transactionAinto blockN+3, upon blockN+2CensorPoolagain censors transactionAby invalidating blockN+3and instead mining blockN+3b, again without transactionA, uponN+2, and another block