It could well be that CoVid19 led to a 50% productivity loss (processes had to be completely reengineered or stood up from scratch without setting foot in the office).
Then, if the actual measured loss is 20% then perhaps it actually went up by 30% (the difference).
One cannot really conclude much from this study due to a lack of control, and I hope this is discussed honestly in the paper.
One thing I could observe is that working from home required a lot of people that are not used to it to learn to communicate professionally in writing, which they weren't used to. With the option of clarifying via a quick personal chat on the corridor taken due to an "all remote" forced work mode, poor communicators really do need a lot of overhead in terms of extra Teams/Zoom/Slack... meetings to get their message across.
Another observation was there was suddenly a lot of extra work. Initially I expected to have more time due to less business. In reality, CoVid19 also presented new opportunities, so extra projects were launched that would not have been required without the pandemic. Another variable that can't really be controlled.
Working from home as a planned/rolled out thing for a percentage of people is vastly different from a pandemic improptu shutdown. I was actually surprised how quickly the whole business world adjusted. I've been arguing for years in favour of working from home but was told it can't be done, or people won't do enough work.