You're right. Sell pressure (capital outflow) cannot end. That said, with inflation (or emission) becoming less and less over time, there won't be as much to sell.
I'm sorry, but difficulty is also security. Attacks aren't just about hashpower, it is also expense (capital to buy and run enough hash) for the attack. If the hashpower goes down, you're right... there is a window where attacks become possible before the difficulty adjusts (I'd argue ETH PoW per block adjustments were more secure), but that would require the same amount of lost hashpower to also be 'found' and dedicated towards attacking the network at the same time. It isn't like the network was less secure 1 month ago than it is today.
Block difficulty does not determine security in any way. All it determines is average block time, that's it. The security of a network is determined solely by the hash power on the network (and the expense to double it). There's no window while difficulty adjusts, if hash power goes down the security goes down until hash power goes up again. The network was less secure if the hash rate was lower. On September 14 at 12:00 UTC, bitcoin's hash rate was 136 EH/s. Right now it is just under 250 EH/s. Bitcoin is more secure right now than it was on September 14 at 12:00 UTC, and the block difficulty has nothing to do with that whatsoever.
The hash difficulty adjusting automatically keeps the block time relatively constant. Security goes down the less hash power is on the network, that's an undisputed fact. Hash power is proportional to energy consumption. Lower energy consumption means less hash power means less security. Think of it like this: if 2 CPUs secured bitcoin, it is trivial to attack. How much hash power and energy an attacker has to buy for a successful attack is the security of the network.
You don't have to factor in the distinction between emission (commonly called inflation) and fees when discussing this particular topic, any bitcoin liquidated to pay for energy usage is mathematically identical to price inflation when looking at BTC as a currency. It is sell pressure, it is capital outflow.
Emission of new bitcoin, or inflation as it is commonly called, is not the same as capital outflow. Capital outflow is sell pressure. You cannot get to 0 capital outflow as long as miners must pay for power. Do you understand what I'm saying? This has nothing to do with how many new bitcoins are created.
Please try to understand what I'm telling you. We aren't talking about the usual everybody-has-heard-it-a-million-times bitcoin critiques. I am very, very familiar with how bitcoin works and understand very much all the points you're making. But these things have no bearing on what I or the article are saying.