this has already been discussed
https://bitcointalk.org/index.php?topic=5502446.msg64306383#msg64306383anyway lets address the points
the first ignorance of the OP's quote is that the assumption that only ~3000 transactions should occur and each transacter should pay more.. when reality is more transactions could and should go into a block and thus each transacter doesnt need to pay more due to more transactions getting into a block to accumulate a better total for a mining pool
the second ignorance of the op is that miners(asics) do not select the transactions. they just hash no matter if a block is full or empty. its the mining POOLS that select the transactions and even if there were not many unconfirmed transactions waiting, a pool can set a minimum fee to accept thus there is also not a need for network policy to have set defaults to bump/force a fee up.. the mining pools make the choices
pools can also decide how full a block needs to be, if the blocks had a network policy of say 16mb. this does not force pools to fill a block to 16mb if it caused network propagation congestion. the pools will see the delay of their block candidate propagation if it got too full and so would decrease transactions in a block even if the limit was higher(as the case of blocks only filled <0.5mb when the limit was 1mb 2009-2013)
so with all that said. the pool can choose the amount of data to include to ensure the data is not going to cause delays in the fastest broadcast competition. we do not need devs playing politics forcing economics on users to force users to pay more when the network can self regulate.
different pools can choose their transactions and if a miner on a certain pool is only getting a few sats for their time due to having too many other miners on the same pool cutting into its reward share, they can move to a less competitive pool.
also when it comes to the cost/reward for miners. they can jump to different pools/coins if they dont get the results they like or temporarily move to altcoins of same hash algo if the market conversion rate of coin to fiat is on a dip. but its worth noting even if a block only has <4btc reward total($280k) its not actually a <$280k cost to break/harm/undo/reorganise a block or the chain.. miners costs are not per block. they are spread out over a ~2 year life cycle of the hardware. meaning ~105000 blocks so the actual security/cost to be malicious at 51% is
$280k * 105k * 51%=~$15b of hardware and electric on average in recent times.. and so if its going to cost $15b to accumulate the hardware and pay electric contracts to mine at these levels. they need to then want to be malicious to a point of getting ROI of more then $15b to break even/profit
most transactions are not worth $15b and even if they re-orged a block to empty block each block for 2 years. unless they owned all utxo's of all them blocks to undo their confirmations or mess around.. they wont break even. so its not worth breaking bitcoin.. even if the per block cost is only $280k of security