Block size is in a sense still increased, well that depends on your definition. Blocks are definitely bigger than 1MB right now.
The most important period to look at would be the block size debate a few years back. One camp thinks that a hard and fast block size increase is the solution.Then again, how much should we really increase the block size? Can it ever be sustainable? The point about a unsustainable block size increase would be that we would reach a limit where only large entities can run Bitcoin nodes, delay across transaction propagation, etc. Big blocks takes longer to propagate, and stale rates would increase. I argue that this point is largely invalid, if you consider propagation speeds of the modern network as well as the inter connectivity of miners.
The other thinks that having second layer scaling as a solution would be the way going forward. Having lightning network was the key thing that made people think that the adoption would skyrocket once adopted and thereby reducing congestion on layer one. Well, it hasn't been extremely effective though adoption is definitely going up.
There are tons of supporters of both sides and they all have their concerns and their vision to what Bitcoin really is. If you want the actual answer, then it boils down a lot more to politics rather than what we can or cannot do.
There should have been a 4-year upgrade in terms of block size incrimination in a similar fashion to how network difficulty rises with time. What I mean is, that Satoshi was able to identify the supply and demand issue earlier through its halving event. However, they did not consider the scaling of block sizes or maybe they never thought that Bitcoin would go to such a high level anytime soon.
So I am not sure if this is technically valid or not, but there should have been an algorithm that automatically increases the size of the block every 4 years or whatever period is feasible to make it happen. Let's say that would be a block-size event in a similar way we have a Halving event. As a result over time, we don't need to fork Bitcoin and create more shit coins as a result.
I know I may not be close to the solution because there could be whole new realm of coding behind this, but at some point we definitely need to think about the solution whether its hardest or simplest, that wont matter.