The specter of changing the block-time refuses to go away. Below are 3 arguments why BCH switching to 1 minute blocks would be very bad. This is not necessarily an argument against 1 minute blocks in general, just against BCH switching at this point. Please steel-man them.
At the point of writing this Antpool controls 54% of the BCH hashrate. 1 minute blocks would increase this, possibly to 100%.
Decreasing the block-time would increase the orphan rate. The cost of an increased orphan rate would be to the detriment of smaller miners, chiefly because a pool with 51%+ of the hash has zero incentive to choose another miners block in the face of a close tie, because they will always end up with the longest chain due to their majority hash.
So since smaller miners are now carrying 100% the cost of the increased orphan rate and the larger miner carries zero, it becomes more viable for smaller miners to direct their hash elsewhere and for the majority miner to continue with the captured chain. You end up with a single mining pool mining 100% of blocks.
You could argue that shorter blocks ‘smooth’ the mining reward distribution which should help smaller miners. This is true, but this alone does not negate the above argument. Smaller miners still bear the brunt of the orphan rate.
Alternatively you could point to other developments in block propagation which could mitigate the orphan rate (CTOR, Graphine, etc). If implemented, these may mitigate the issue, but that’s a big if. Also, they still will not negate this issue completely.
Block-time is too fundamental a thing to change at this point. Anyone who made a last will and testament style tx to activate at a particular block height would see these broken. Even if no such tx currently exist then changing something so fundamental at this point would deter anyone from ever making such a tx on BCH. You cannot build your castle on ever shifting sands.
Let us be honest with ourselves. Too controversial a change will lead to a fork war. Nobody wants that at this point, but we must be realistic. While the ability to fork is what makes a project truly permissionless, even the mere prospect of yet another fork might prove too exhausting for many in the community.