So, you're saying that block production follows a Poisson distribution and there's no technical explanation for it? I've looked into this distribution, and it makes sense because it's suitable for modeling the number of events (block production) that occur in a fixed time interval (10 minutes, difficulty adjustment) when those events occur at a constant average rate and independently of the time since the last event.