The fact that the relation is one of an amendment or a reconciliation to new data is a point in favor of skepticism against his original assumptions, but not a defeater.
I would agree that we can credit intuitions as being more than mere assumptions, and I think you're right about The legitimacy of positing assumptions as long as they're posited in a spirit of openness and responsiveness to philosophical and scientific criticism. What do you know, a healthy and reasonable of reply on philosophy. To tie this to Daniel Dennett, I have found one of his most valuable contributions to be a notion of the principle of charity, which is to credit the person you correspond with with the most reasonable possible interpretation, and I think the best philosophical exchanges are carried out in that spirit, and so I think he would highlight yours as the response in this thread that best embodies that principle.
My understanding is that post-relativity, Neokantians have been trying to reconcile Kant's thought with the implications of relativity. Philosophy as a discipline is, as I understand it, happy to work with new scientific discoveries, and not going to stick with philosophical dogma in the face of contradicting science. Besides, beyond a certain level, doesn't science have to give way to philosophy? I don't see holding certain philosophical assumptions (not really assumptions, but intuitions or propositions argued from intuitions) as a necessarily bad thing or even an avoidable thing, as long as it's open to both philosophical and scientific criticism, where applicable.