They're just looking at the trendlines.
Follow me here -- 81% of these people were a net drain on your company.
And peak productivity here would be if you never even hired them.
So working backwards, and assuming there is a correlation here, if you are most productive when you reduce their hours 100% ... you should see at least some productivity gain if you reduce their hours 20%.
So they're right -- but you've got to look at the big picture, don't accept the small boost a 4 day week would give you, go for the big boost by getting rid of them entirely. ;)
All that said, for a lot of jobs a 4 day week works out fine in practice. A company I work with went to a 4 day week, and saw no loss in productivity. They manufacture stuff - and rotating the staff around the week worked fine, the machines still run all week. For other jobs it won't work out.
Arguably instead of a 4 day week, they could have just cut people instead, and maybe even gotten the same productivity from fewer people, (and boosted profit), but maybe not -- fewer people working that much harder would lead to more burnout, more mistakes, higher staff turnover (leading to lower production and more mistakes again), plus higher training costs. Management didn't think that was a winning move.