no one still have [...] explained [...] why simply increase the block size MB do not scale, unless "scale" doesn't simply mean having more transaction per second
Carlton Banks did explain to my satisfaction what he meant by it. If I understood him correctly, another way I'd describe it is if you change the scale of a map you get a lot more area shown for each increment increased. According to Carlton Banks's argument, this means at best, we get 'only' double transaction volume by doubling the block chain. For (I'll avoid the word scaling) increasing the potential transaction throughput to a number that would satisfy potential demand, we need something that does more than linear growth.
From the other heated argument above I can see there may be a reasonable case to be made that you don't even get 1-1 (throughput / blocksize) which if true, makes it even more important to find longer-term solutions - of which segwit may be one.
However, not everything has to be addressed at once and to me - as to Gavin and others - 'kicking the can down the road' by having an increase that would have prevented us from getting to what we have today - a currency dropping off on-chain use cases by the day. To this day I fail to see how it can be worth risking the totally unknown consequences of the presently worsening situation - deliberately letting it get to the stage of a ceiling-hit-forced-fee-market - in order to 'encourage' or speed up the development of second-tier solutions.
Talking of Gavin, I omitted your references to 'flaming and pretentious' and dismiss the oft-repeated 'Gavin [having gone] full retard' as an example of some of the phenomena I talk about in the last post. I'm trying to put such statements to one side and not let it colour my preparedness to listen to the reasoned arguments people who are saying such things are making. At the risk of sounding patronising, I suggest others participating here do likewise.