No, PubMed usually just posts the abstracts. The way they do things, we pay the taxes, the taxes go to, say, NIH in the form of budget dollars. NIH may give the money to a sub-group of their own, which funds a grant to some foundation on non-profit to do the actual work. When the work is completed, the authors can allow one of the journals to publish it. And THAT is the point at which they will edit (sometimes only slightly - but you can bet there will be something) for publication, and then they demand copyright protections. Because they are NOT direct government employees, they get away with it., and with the claim that what those government dollars produced is no longer the same work. And so, even if work done by direct federal employees is public domain, you will not even know that it exists - until or unless that abstract appears. And it will still be locked up behind that journal's paywall, unless the actual report is government published, directly. Your only access would be through a FOIA request, which could take many months, if you ever get anything at all.
That's the existing game. So the question is - how far will this go, and will it fix that particular problem?
Edit to add: If you doubt, wander on over to PubMed and just see how many full articles you can actually download. On the flip - those government offices typically hold subscriptions services to all those journals, so if you happen to work there, you can read whatever you like, pretty much. Try that on your own from your home computer, and see what you get! (It won't be much.)