You know, this is a little interesting. Nanopool has a broken rpc implementation which causes issues when e.g. submitting two or more shares in quick succession. Now, we assumed that we'd still get replies for all submitted shares, but there could cases where they actually drop rpc requests. I didn't do the fix though, need to check with todxx.

Comparing pool payouts is a different thing though. Unless you're sure the pools hit exactly 100% avg luck over the same period, it doesn't say much. The poolside hashrate displayed in the miner should match your miner hashrate minus dev fee and any rejected shares. If that looks good, we can't really do much more from a miner dev perspective. If that number is ok, but the pool reports a different hashrate, something is fishy.