We see a savings of $400 million and think "we should do this!" But it's a drop in the bucket even if it were that much of a savings.
If each government employee needs to change their font, or needs to set it as the default font, or needs technical support to configure the defaults in their word processor. If IT needs to modify images to use this font as a default. Just these actions are going to cost a significant portion of that $400 million when you consider it across the millions of federal staff.
This also assumes things like the government is actually paying for ink or toner in quantity, instead of, for instance, holding a contract with Xerox who charges per impression rather than based on how much ink you use.
It also assumes that there is no difference in legibility between the fonts. That people with vision impairments will not have difficulty with reading the document.
An easy way to think about whether an initiative like this is reasonable is to think about whether it makes a lot of sense for any individual to do. Do you think you, individually, could realize any significant savings by changing your fonts? If it only makes sense when millions of people do it at once, and even then only when certain assumptions are met, and then only saves a few dollars per person per year, then it actually is more likely to cost a lot more in overhead to make sure it happens than it will ever save.