Interesting if you look at the spreadsheet where the graphs come from, weekday car traffic is consistently 92-94% of pre-covid, with 100% on weekends. So it seems reasonable that WFH is reducing overall road journeys by about 7% at most. Although fuel prices going up 30% may have had some effect.
National rail figures are 7-day rolling averages so you can't compare weekday/weekend figures and I haven't seen them elsewhere, but I'd assume a higher proportion of rail trips are for commuting journeys thefefore I'd expect the drop in commuting passenger numbers to be >7%. The overall numbers are bouncing around between 85-90% for the last few weeks.
The tube figures are roughtly 75-78% weekday and 90-95% at weekends.
Having a look at the National Travel Survey stats for 2019, indicates 20% of car trips were commuting, 47% surface rail trips were commuting, and 60% for tube travel.
If the 7% drop in weekday traffic was solely from a reduction in commuting, this would be a 35% reduction in commuting trips, which is higher than I'd expect across the population. But applying the 35% drop to railand tube commuting journeys would give an expected (0.35*47 = 16 drop )84% of pre-covid passenger numbers for surface rail and (.35*60 = 21% drop) 79% for tube, which isn't too far off.
For surface rail, 84% on weekdays and say 105% on weekends would give 90% overall which seems to be not too far off where we are. I'd guess the drop in commuters might be

5% but business travel will also have dropped, possibly by more, which was about 10% of rail journeys in the 2019 survey figures so that would knock a few % off. Weekend figures might also be >105% given the various anecdotes I've read, but I've not seen any actual figures.