[frankenst1]

> Is the energy cost per transaction in real world for bitcoin going up or down? Everything else is trying to muddy it up from my personal perspective.

Energy is expended to secure the network as a whole and not for the transfer of coins.

But if you insist on computing such a metric with only limited applicability, the answer is: Yes, the metric goes up:

Estimated power consumption per day, 2020-2021 +5%:

    01/01/2020: 0.2 TWh
    01/01/2021: 0.211 TWh

Estimated number of onchain transactions per day, 2020-2021: +6%:

    01/01/2020: 288K
    01/01/2021: 305K

And this doesn't even include all the hidden private transactions that occur off-chain on Layer 2 solutions like the Lightning Network:

Number of nodes on the Lightning Network, 2020-2021 +60%:

    01/01/2020: 5K
    01/01/2021: 8K

Capacity of the Lightning Network, 2020-2021 +430%:

    01/01/2020: $6M
    01/01/2021: $32M

[frankenst1]

> Yes, the metric goes up:

Oops, meant to say transactions go up vs energy, so the metric energy/transaction goes actually down.