Thank you for taking my hand and guiding me through implemention choices. 🫂 Much value from you to me.
Your sidenote is acutally about my main point of my original note:
**Theoretical**
The increased sats/vb rate (=delta sats to pay for completely new data to be distributed in the mempools) is for the signer from mempooltx[t0] and mempooltx[t1] and mempooltx[txyz] very cheap.
**Practical**
I then can distribute for just a few hundred sats completely new data with an RBF "overwrite" of the old data in the mempools.