I wrote up a big post on this a couple of years ago. The arguments are old and tired and fail to see the bigger picture:
My point #9 I think is the one that laypeople understand the most:
  1. bit of a gripe: “That’s about 70.4 TWh in a year. Which is about as much energy as Colombia or Bangladesh use.” — I find the constant usage of “as much as X country” as a comparison to be alarmist, headline-journalism, marketing buzz. US clothing dryers alone consume ~72 TWh/year (US Residential power 2019 (1.44 Trillion kWh): https://www.eia.gov/energyexplained/electricity/use-of-electricity.php and Breakdown of Residential Energy [5% clothing dryers]: https://www.eia.gov/todayinenergy/detail.php?id=37813). Not to say “What about dryers” as a whataboutism, but this “as country” framing bothers me because I see it as a bad alternative to comparing like things. Bitcoin is more like a clothing dryer than it is like a country. The author at a later point brings up GDP of countries as if it’s a valid argument for energy usage (Bitcoin doesn’t have an equivalent GDP of the country it matches in energy usage) but this is a misleading concept as Bitcoin is a technical utility, not a country. Clothing dryers also don’t have a GDP. I don’t see articles saying that other globally used technologies and practices use the energy equivalent of a small country, even though they do, which leads me to think of this framework as propaganda every time I see a headline comparing bitcoin energy to a country. Comparing Bitcoin to other global technologies and systems like clothing dryers, idle electronics, etc makes it more relatable and accurate than comparing it to reports of energy consumption by entire countries with all the complex reasons they use electricity (including mining). Energy usage data by countries aggregated across all reasons is easy to find, but estimating individual energy usage categories across the globe requires the same level of research as estimating bitcoin consumption — so I get why it’s not done (because it’s work), but it also feels like we aren’t doing the full job of research by just grabbing at some data that’s easily available and turning it into a series of headlines.

People also miss that Bitcoin can run with only a couple of raspberry pis and process all the transactions and mine all the bitcoin. The hashrate is an economic effect of the demand for competition in making blocks, not a measurement that can be used to evaluate the cost of a single transactions by dividing the power consumption by use.