Correct. This is the same reason why Proof-of-Stake (PoS) cannot work as people hope. Due to fundamental economics, people will always spend according to the value of the new currency issued by mining. Switching to PoS just means they will spend in some other way in order to get those new coins.

This also means that as the Bitcoin mining reward diminishes, the money "wasted" on energy is going to decrease, unless the transaction fees rise enough to match what the reward was paying. In that case, because people are willing to pay those fees, it seems quite logical that the cost of the mining is providing a value equal to its cost. Otherwise, people would not be paying those fees.

I really don't know if the Bitcoin experiment is going to work but a lot of criticisms of it are based on some really poor logic.

That is not true.

The upper bound is the combination of how much people will pay you for the bitcoins they mine, how much people value being able to transact (transaction fees), and the value of existing held bitcoin.

Since mining is what prevents attackers from stealing bitcoin,the total value of all bitcoin in existence actually has a bearing in this.

If no one were transacting and no coins were rewarded for mining, people would still mine just so that their existing coins remain secure.