I also think decreasing the difficulty won't have the same effect because like what franky1 said it will speed things up and will ruin the countdown of the block reward halving.
If you wanted to reduce the block time target, let's say, to 1 minute, then you obviously can do the following:
- replace DifficultyTarget by DifficultyTarget/10
- replace BlockReward by BlockReward/10
- replace NextHalvingBlockHeight by HardForkBlockHeight + ((OldNextHalvingBlockHeight - HardForkBlockHeight) * 10)
and you will have, approximately, the same halving schedule and the exactly same coin supply.
The reason why block times are not reduced is actually, as some have already said, the higher orphan rate, that would lead to a less efficient system. But there is some research done in this topic, so it's not sure it won't be implemented at some point in the future if orphan rate can be decreased by technical improvements.
Well, the orphan rate plus the unfortunate potential consequence of the total overall blockchain growing rapidly in size. A fair assumption is that the average block would be a bit smaller if it were faster because less transactions would be queuing, but if more spam got included and we (hypothetically) averaged 0.3 or 0.4 MB blocks every minute, that's a pretty hefty size increase over time compared to a maximum of 1MB every 10 minutes. You could offset this by also reducing the max blocksize, but then it doesn't allow for more transactions.
and use LN to make BTC into an Offchain Fractional Reserve banking system.
How would that be possible? AFAIK LN would not allow "fractional reserve" and that's the advantage it has over centralized off-chain solutions like Coinbase.
It's noble of you to try, but kiklo isn't listening to reason on that front. Believe me, I've made multiple attempts to explain why he's wrong, but he keeps repeating it regardless. LN is fractional reserve, up is down and the sky is green with purple spots. That's just how things are in kiklo's world, heh.