1. Those estimates are way off. I don't know how they get those numbers, but clearly we need more accurate tools for fee estimates.
  2. The estimate for "≤ 20 min" is always going to be extremely high from time to time. 13.5% of the time there won't even be a new block in 20 minutes. When on-chain fees are a concern, 24 hours or more would be a more realistic target. Yes, having to choose between very slow (and still rather expensive) and very expensive (and still potentially slow) limits what activities can be done economically on chain, something that should have been clear since fifteen years ago.
reply