You mean fixing on-chain scalability? Increasing the size of the blocks was also proposed before by Roger and CSW group but it was rejected. Instead, Segwit was implemented to help improve network scalability and reducing fees.
Segwit is in effect a block size increase in a manner that ensures older clients are still functional.
I'm not a developer (not even close) but I understand that improving scalability of a blockchain without compromising its security and decentralization will be difficult.
Block size is a tradeoff between security and scalability, with the security factor decreasing as time goes by due to the far more efficient network. It is incorrect to say that we will be far worse off in terms of security and decentralization given a block size increase.
If you want Bitcoin to scale, you need a block size increase in conjunction with a 2nd layer payment network. You have to initiate the channels, in LN with an on-chain transaction and an on-chain transaction to close the channel as well. You cannot accommodate for those without having to increase the block size or make transactions more efficient.
The problem with solely increasing block size alone is that it is not a panacea for the scalability problem. You cannot increase the block size to compete with the next best payment processor. There is a limit till block size becomes excessively large and it won't be feasible to do so then. ECDSA signatures has to occupy that much space in a transaction while data compression is possible, I doubt it would be able to truncate the data to that extent.