You can’t do 1 second blocks times on POW because you will get tons and tons of reorgs. Basically some pool might find a block at the same time as another but they are delayed in sending the hash to the network and the other block can get accepted while they find another block right after and the block prior gets reorged with the higher block number. Hence why it’s not good having very low block times.

There is also issues of space and bandwidth. Many have already tried this and failed. Like BSV with their huge 1Gb block (or some other high number can’t remember). Nobody would want to run a node.