Confirming on the testcase, Win XP.
Does this require jquery to reproduce?
Disabled state is persisted for buttons across a soft refresh, along with other form control state. I believe there are existing bugs on that.
Is it intentional?
We have code to make it explicitly happen, so yes. The big question is whether it's desired. And the problem is that that depends on the use cases... If the button is disabled in response to the values of other inputs on the page, for example, then restoring those values should also restore the disabled state.
It's not consistent with other browsers. And if you ask me, the coder (JS) should take care of these things, not the browser itself. E.g. run validations on DOM load. There is just too many cases where this is not the desired behavior.
Dup Bug 654072