Not really. The network spends those kWh whether you send your transaction or not.
It's also worth to note that nobody really knows how many kW miners really burn. All the figures are more or less educated guesses from the efficiency of equipment sold publicly. If someone has a more efficient method they probably won't share it.
But nobody accused Bitcoin of being efficient, ever. That's what almost everyone notices when reading about it first time.
The cost is per transaction, not per amount transferred. A small transaction takes just as much space in the ledger as a large transaction.
A fun thought experiment is to see just how limited the throughput of Bitcoin is. Suppose the Bitcoin enthusiasts were to get their wish and Bitcoin becomes the dominant "currency". If every person on the planet were to make a bitcoin transaction once if their lifetime, that alone would be too many transactions/second for the network to handle.
Pretty much. It is only useful in cases where you are making very large transactions. At that point, you have few enough actors that setting up a trusted network is far, far cheaper and more flexible. The trustless nature of Bitcoin is what makes it unscalable.
Its not correct. The network spends this energy either ways, and it not only allows transactions to happen but it is integral to securing funds that are not moving. Dividing the energy consumption by number of transactions is therefore nonsense.
Depends on the transaction fees associated with those transfers. If those millions of people also post higher-than-average transaction fees, then the Bitcoin network prioritizes those payments, and cannot process any other transfers.
Nobody will pay a coffee with BTC, just like you don't pay for your stuff with a gold coins.
Some say Litecoin will be used for that but I doubt it. Coins like Cardano or IOTA will be used for transactions. I'm not sure about Cardano but with IOTA you can transfer a fraction of a cent (or no value at all) without fees (that doesn't mean everything will be always fee-less).
Analyses like this just take the energy cost of each block of transactions and divide it by the amount of transactions in the block. If the number of transactions in a block doubles, the average energy cost per transaction is halved.
Another way of stating it is "the marginal energy usage of a bitcoin transaction is essentially 0".
Except that there is a maximum number of transactions per block. Across the entire network, by design, there can never be more than 5 transactions per second. This is stupidly small. If people received their biweekly paychecks in bitcoin, only 6 million people could be paid without going over that transaction limit, assuming that absolutely nothing else is done using bitcoin.
The marginal energy usage being zero is another way of saying that Bitcoin wastes the tremendous amount of energy that it does even if nobody is using it at the time.
Yes, but the number of transactions in 'bitcoin' as it's currently defined is severely limited (blocks are already full at a $15 per transaction cost). Conceptually, making something which increases the transaction count involves forking bitcoin, and a fork which aimed to do exactly that was denounced heavily by the community and rejected by the markets. The marginal cost may be small but the cost per transaction is high, by design, and by design which has extremely heavy resistance to even small and simple changes.
