I disagree on the "it won't happen" part, because, well...maths... It's not a matter of if. Only a matter of when. It could be 30 years or 60 years, but at some point even 1GB or 1TB blocks will be peanuts for the standards that exist in future time-space coordinates.
I agree on the "we will never be able to process more than a centralized...", because, well.... p2p inefficiency.
These two statements make zero sense due to the fact that they are contradicting themselves. I have only claimed that Bitcoin can not catch up to Visa with on chain transactions due to the infficient way of scaling decentralized systems (in comparison to centralized). I have no idea what kind of math you're talking about here nor how it is relevant to this statement.
You obviously have little knowledge on the scaling of large systems (distributed vs centralized vs decentralized). I'm not saying that we will never be able to process a lot of transactions; I'm saying that we will never be able to process more than a centralized and well expanded system like Visa (on chain transactions).
Since Bitcoin, in your opinion, can't scale well, we should forget about scaling altogether?
Why bother with segwit then? Why not go for *smaller* blocks & foster "decentralization"

Apparently you are unable to read the fine print. Segwit -> IBLT & weak blocks -> LN -> possibly some smaller block size increases and we should be fine for quite some time. LN is a wonder when it comes to 'tps' which becomes actually irrelevant with it (number of channels will matter). The main idea behind Segwit is actually not the increase of the transaction capacity. If you had done enough research you would know that.