That's not a very good description of what Luke's (latest) Hard Fork plan actually is; the blocksize limit depends on absolute blockheights in his proposal, and so whether or not the blocks are bigger than 1MB or smaller depends on when the fork is activated.
If that hard fork activated tomorrow, we'd have 300 KB blocks. But that's very unlikely to actually happen. By the time a fork proposal like that activated, it could easily have reached the stage of progress where the blocksize is bigger than 1MB (I think Luke is using the 17% per year "technological growth" factor that Pieter Wuille did in BIP103)
I think it's a fair description of what it is. It initially lowers the block size, which is exactly what I wrote. Take a look here:
if in 2018 January, it would begin at ~356k; or if in 2024 June, it would begin at just over 1 MB.
Block size is only going above 1 MB in 2024 with his proposal.[1]
Yes, and under that proposal, the 1MB limit can remain until Lightning begins to soak up some transactions. You can't make a real case for 2MB until the true capacity of operational Lightning is observed in the real world, I think that's reasonable.
Why increase the 1MB size if blocks became more empty than they are now after Lightning is available? I think most sensible people in cryptocurrencies recognise that full blocks are an important requirement for getting the miner/user incentives properly aligned for establishing fees.