+1 for your very interesting question, while most of the community is very optimistic about the future of ETH 2.0 despite some problems happened in the Testnet.
As you can see, transaction fees are pushed up by the limitation of the network's scalability compared to the huge demand of the community, thanks to DeFi

, so to reduce transaction costs then we need to calm investors down and increase the service speed of the Ethereum network, from 15tps to 15000 tps, or at least 150 tps

In addition to solutions developed by a number of supporting projects, one of the solutions is expected to have a positive impact on the scaling problem - Shardings. I hope Shardings will bring a good result for the entire Ethereum community, 1000-3000 tps is more than enough for us to be satisfied!
If the speed of the network does not increase, or the increase is not dramatic enough while the demand for transactions continues to soar, if transaction costs are still too expensive as they are, Ethereum will not die, at least according to my judgment. ETH 2.0 using POS will solve the security + mining problem, transaction speed can be the object of the update to ETH 3.0
