I guess Luke likes to shock others, he himself couldn't seriously expect such proposal to be accepted. Nowhere in the agreement it is stated when actual blocksize increase must happen. He exploited this uncertainty to pervert the meaning just for lulz. If we are playing such games, soon we will need teams of lawyers and our agreements will become tens of pages in thickness.
He did not just "exploited this uncertainty to pervert the meaning just for lulz". If you have read any of what Luke-Jr has said about block sizes and capacity, he genuinely believes that the block size is currently too large because it is difficult to start running a new full node. It is also becoming unsustainable to continue to run a full node since it costs so much in bandwidth, processing power, RAM, and disk space. It wasn't proposed "just for lulz" or to troll anyone but rather because Luke-Jr genuinely thinks that decreasing the block size would help Bitcoin by allowing more people to run full nodes.
On the other hand Sergio's proposal follows the spirit of the agreement. It is simple, honest, it gives each side what they want.
It does not give both sides what they want. Many who support segwit clearly don't support the proposal. A lot of that is because they think that having a 2 MB base block size is too large (that means that witness block sizes can go up to 8 MB). Furthermore, that still allows a 2 MB legacy transaction which takes a long time to validate due to the quadratic sighashing issue. Additionally, since it is a hard fork, there are a number of other things that should be included into it as we don't want to have multiple hard forks but rather one with everything that we want that needs hard forking. Segwit2MB does not address that.