For a bit of fun, I created a low fee TXN (493 s/kB) to see if pool would confirm below 1000 s/kB. Seems the answer is usually “very very unlikely”. My economic game theory went something like this…
Given the option to mine no TXNs or to mine TXNs below 1000 s/kB, miners would always choose anything over nothing
The cost of including a TXN, is practical nothing in comparison to the cost of hashing a block. The TXNs are just included in the root of the block header, so not really a heavy expense. Perhaps I’m overthinking it. Maybe most miners just use the default settings and don’t waste a moments thought on anything beyond it. Maybe there is an economic factor I’m missing.
Ohh well. At least PR #13990 is staying active… Guess I’m just impatient.