Interesting take. Although I hate to be a wet blanket you are omitting an essential aspect of your speculation: economics. Say that your estimate is correct and over the course of a mere 2 years beginning in 2021 global oil use declines by 25% and Exxon among many other large oil companies (and all of the related servicing companies etc.) were so short sighted that they are forced to declare bankruptcy along with a large part of the current car industry (dealerships, parts suppliers, auto shops etc.) the resultant chaos in the global economy and massive unemployment this would cause would drive the planet into a brutal recession. Energy in it’s current form accounts for 9% of US GDP and slightly less globally. Autos employ approximately 3 million people directly in the US alone and millions more indirectly. Of course some will end up finding work among the EAV companies but given the numbers you are putting forth in this article perhaps 75% (that’s probably conservative…) will become unemployed over just 24 months. Oil and gas supports 2 to 3 million directly and closer to 10 million jobs in the US indirectly. If you seriously think that over just 2 years we could lose 75% of the input from these two industries both in terms of GDP and employment, then the article badly misses the point. The carmageddon you are referring to, should it happen as you are predicting will trigger the worst global recession in history. I seriously doubt anyone will be wearing Alexa spectacles should this occur.
P.s should the cost of owning and operating said cars fall as dramatically as you predict then the number of cars on the road per Elon Musk (see the boring company) would rise substantially not fall. The necessity for the idea of mobility as a service to be successful is that the availability/convenience is at least equal to that of having a personal vehicle. For this to be true these cars would need to be everywhere, all of the time. Given that roughly 50% of metropolitan workers live in the suburbs and commute to and from work everyday you would require at least as many cars on the road as there are today just to service the rush hour commutes.
P.p.s this analysis and many more like it ignore a fundamental aspect of having a personal vehicle: you own it. To believe that attitudes towards travelling in a public vehicle with a group of strangers will suddenly change is silly. If you assume conversely that this is not the case and that each traveller will have their own public use vehicle to travel to and from work in then you will still need as many cars and all of the roads as you do today. No matter how I look at it I fail to believe that suddenly the vast majority of people will be fine with waiting for a vehicle to show up in the morning, riding in it with strangers to wherever, figuring out who gets dropped off first, then repeating this terrible exercise at the end of a long day. It just ignores too many obvious human preferences. Perhaps you could convince these people to ride alone but shared seems very unlikely to me for a large portion of those that make this journey every day.
