...
I think we'll reach a compromise where a load will go back full time, some will do mix like 3-4 days in and 1-2 days out, or alternate things and a small elite few WFH majorly.
Realistically we might not know that until everything does shake down; however, just as one data point, my employer (not a very large one, but still) has told us that post-covid, we're all allowed to WFH part time, and 2 days per week in the office is acceptable. (This isn't just idle talk - new contracts have been issued, so they aren't going to walk that back easily!). Talking to my colleagues, not all of them are planning to WFH to the maximum allowed - which backs up your points 1/2 partly -
but of those of us who have a decent length commute,
we are all planning to do so. It's the people who live within walking (or short cycling) distance of the office who'll do more days in there!
(This makes a lot of sense - if "going into the office" means "less than a mile's walk", then sure, it's not much imposition. If it involves more time
and cost, well, of course you're keener to avoid the commute and do WFH!)
The end result is that, of the people in my office who
would otherwise travel by public transport, we're all going to continue part-time WFH to the extent we're allowed to. (It's also the natural response to your point 3 - my employer doesn't need us on site
all the time, but will need us on site - or at a customer site! -
occasionally.)
This sounds a lot like what Nationwide (who are a large employer!) are doing - they're not mandating "everybody WFH 100% of the time", but they're allowing a mix, and a lot of people want to do so at least part of the time.
Will it cut commuting by 90%? No, not that much; some firms just won't or can't allow WFH, even among firms that do, some people won't want to WFH, as you indicate. But even a 20% drop in commuting is a large enough shift to make a very noticeable difference - most businesses couldn't ignore that many customers disappearing overnight!
(Something that I'm sure is being considered: when I started my current job, I bought a season ticket. Once you
have a season ticket, there's a strong incentive to
use it - you've already paid for it! So you use it for leisure travel (possibly buying ticket(s) for your kids), on a work day even if the weather's foul and you don't fancy the walk to the station, well, I'd rather not drive, I've already basically paid for the train ... Now I won't be buying a season ticket, those incentives are gone - I'm not "tied in" to taking the train, and I can skip it freely any day or week without feeling I've lost out. How big an effect is that? I don't know; maybe not big at all! But I expect there
is somebody whose job it is to figure out how much additional revenue you might lose when somebody stops travelling by season ticket in favour of individual day tickets.)