My suggestion was to introduce a minimum fee in the Bitcoin system. And it is better to introduce such changes at the protocol level. You claim that the disadvantage is that "Need to change Bitcoin protocol to explicitly disallow transaction low fees". 
Although this is not entirely true. The minimum fee can be entered at the level of miners, without changing the protocol. If, for example, 95% of miners vote for this and activate this mode, then blocks with transactions with a fee less than the established one will be considered invalid by the rest of the miners. But I emphasize that it is better to introduce such changes at the protocol level.
Although this is not entirely true. The minimum fee can be entered at the level of miners, without changing the protocol. If, for example, 95% of miners vote for this and activate this mode, then blocks with transactions with a fee less than the established one will be considered invalid by the rest of the miners. But I emphasize that it is better to introduce such changes at the protocol level.
Bitcoin network isn't only about miner, but also node. If Bitcoin community don't agree to the change, it'll fail and we'll see 2 chain.
1. Miners can still fill blocks with their own transactions (spam from miners). But so far, there is no evidence that this happened in reality.
I think transaction with 0-fee could serve as evidence, even though the goal isn't to attack Bitcoin, but to prevent fees on transaction they actually want to create.
3. The community has the ability to react to such an event. Most miners should block the "malicious" miner.
If the majority of miners cannot overcome this phenomenon, then the bitcoin community can replace the miners.
If the majority of miners cannot overcome this phenomenon, then the bitcoin community can replace the miners.
Not realistic, i doubt miners would block malicious miners as long as the block itself follow Bitcoin protocol.
Does that mean you found data used for BIP 103? I skimmed https://lists.linuxfoundation.org/pipermail/bitcoin-dev/2015-July/thread.html#9763, but couldn't find any link of data supporting it.
No. You can look at historical data regarding increases in processing power, and network capacity, both at specific price points. You can also look at near term projections of both of the above. I know, but each source might have slightly different number and how they measure the growth.