The more important Bitcoins becomes, the more likely a congestion failure will be absent a scaling solution. Let's assume half the transactions are either not important or urgent. Then what happens if that number grows by a factor of five? This is not unprecedented. There is no way the price will sustainably go up without attracting new users EXPONENTIALLY and the network won't be able to handle them no matter how much in fees they pay.
Let's use a hypothetical example: Imagine that the Chinese government announces that they will soon close the loophole that allows their citizens to avoid capital controls via crypto. This could cause a mad rush to the exits, particularly if the economic conditions inside China deteriorate. If one tenth of one percent of the Chinese try to get their money out through bitcoin, that's (1,600,000,000)X(0.001)=1,600,000 so at only one transaction each, we're looking at over four times the total daily network capacity IN ADDITION TO all the other traffic. That's just one example. I can think of dozens more. Hyperinflation in a third world country. another Cypress type banking crisis. A terrorist attack on the Bank of International Settlements. SWIFT or ACH getting hacked. etc.
The mempool has already been over 10MB for two days straight. We are looking at a permanent and growing backlog of transactions.
that doesn't contradict my point. you are right if a big flush of TXs comes their will be a BIG mempool and essentially alot of TXs would have to switch to another coin. nevertheless its not a "congestion" as this basically paints a picture where NO TXs AT ALL will be processed ands thats bullshit.
if you have 1.6Mio TXs and only 200.000 are fitting on the blockchain then a market will occur and the most important 200.000 will be processed. i know you can argue that that would be not a good situation for Bitcoin OK, but its not that the newtwork stops processing TXs.
i hope it gets clearer now.
Those people switching to another coin may never switch back.