Transaction fees have been about as low as they can possibly be for months. It costs roughly $0.15 to move a UTXO right now.
The people who argued for increasing the block size were making a tradeoff between decentralization and "low" transaction fees. Increasing the block size makes the chain grow faster which causes several problems: the time to sync new blocks goes up, initial chain download time goes way up, and block verification takes more CPU.
The parameters were initially set so that anyone could run a full node for as many years as possible which is the most important part of the ecosystem by far. If you make it harder for the average person to run a node it doesn't matter how cheap the transactions are if ultimately a small group of people are capable of running it. You could argue that doubling or quadrupling the block size doesn't hurt decentralization, and maybe you're right, but that also doesn't move the needle much for transaction throughput either at least not enough to make a difference for adoption.
Bitcoin can and should be scaled in layers. The detractors are just impatient.
While it's not here yet, Vitalik's future plans for Ethereum involve making it possible to run validators on smartwatches[1], which is far more accessibly decentralized than BTC could be and far more possible with Ethereum's PoS consensus as opposed to PoW, without needing to cripple transaction speed to boot.
The people who argued for increasing the block size were making a tradeoff between decentralization and "low" transaction fees. Increasing the block size makes the chain grow faster which causes several problems: the time to sync new blocks goes up, initial chain download time goes way up, and block verification takes more CPU.
The parameters were initially set so that anyone could run a full node for as many years as possible which is the most important part of the ecosystem by far. If you make it harder for the average person to run a node it doesn't matter how cheap the transactions are if ultimately a small group of people are capable of running it. You could argue that doubling or quadrupling the block size doesn't hurt decentralization, and maybe you're right, but that also doesn't move the needle much for transaction throughput either at least not enough to make a difference for adoption.
Bitcoin can and should be scaled in layers. The detractors are just impatient.