Exactly
If half of the workforce decide to work from home just 1 day a week, that's still leads to a major drop in demand for public transport.
A company with 100 employees, 50 of them want to work from home 1 day a week, so every day 10 fewer people are coming into the office
Of those 10 people (this is a London office) 6 use the train, 2 the bus and 2 cycle
Thus rail use from this office drops by 6%, just from letting people work from home 1 day a week
Even a 12% fall would likely only means that rail growth was rolled back by a few years.
However something which is often overlooked is that a reduction in 5 day a week travel would also impact on car use, potentially to the benefit of rail.
Take for example someone who drives 15 miles each way to work and had a total mileage of 8,000 miles a year, if they WFH 3 days a week their annual mileage falls to 8,000 miles.
However some of their personal milage could also fall as they are at home and able to take deliveries (often cited as a reason why people don't like internet shopping). Now if you're decision to drive was because it was a little quicker and a little cheaper then both of those factors become less of a concern.
As even at 4,000 miles chances are the cost of car ownership would be about £2,500 a year. Well if you're rail ticket falls from £1,000 to £700 (due to paying full wack for each of the days you go to work, you could make a bit more of a saving of you worked Thursday Friday one week and then Monday Tuesday or the next before repeating in subsequent weeks and could use a weekly season ticket), however that still leaves your account £1,800 for your other travel.
Given that was about 1,250 miles, and unless you're doing much by taxi chances are you'll be able to manage to do that for less than that. Especially if there's a 100 miles of that (so 2 miles a week) which you walk/cycle instead of drive.
I'm not suggesting that there's going to lots who do that, but it's likely that it would soften any falls we do see or (more likely) help drive future rail growth.
To see a 20% fall from reduced commuting (commuting is about 65% of rail travel) we'd need to see (as an example) 50% of all jobs which people use rail for WFH an average of 3 days a week. That's quite a lot and is why my thoughts are a 10-15% fall would be more likely. As a 10% fall would require 40% of jobs WFH an average of 2 days a week.
Now whilst anyone who did WFH full time wound shift that figure higher you'd only need 3 people WFH one day a week to be back to that average.
Anyway, even those who fuel that they WFH all the time if they go in to the office (or meeting room, or pub, or wherever they have face to face meetings/socials) an average of once a month they are still traveling 5% of the time of someone who's in the office 100% of the time.
Now, whilst that travel is less likely to be at peak hours, actually that's probably a good thing for the rail industry financially. However it could actually mean that they are traveling from further out. For instance if you were only needed to go in for a couple of days every few months, would you rather live in Richmond or live in Devon?
You'd never travel daily in from Devon, but you might travel in on a Monday (doing work on the train) be in the office Tuesday and Wednesday and travel back on the train Thursday (again working on the train) if you did that 4 times a year. However the rest of the year you'll be able to stop work at 5 (having started at 8am) and go for a walk on the moors or along the beach. You may even be able to do so during your lunch break. That's going to be a massive benefit to your quality of life.
Given the cost of housing you could sell your Richmond house buy a London flat and a decent house in the country. In doing so you could work on the train on Friday, spend the weekend in London, work in the office Monday & Tuesday and go back on the train (again working on it) on the Wednesday. Do that 18 times a year and 16% of your working days would be in the office, but you'd still be able to do the sorts of things which are easier to do in London (catching a show, shopping on Oxford Street, going to museums, etc.) but still having a lot of time in the country.
Devon might not be to everyone's taste, so maybe the Peak District, Yorkshire Moors, Lake District, Pembrokeshire, take your pick or add in your own place you love.