I kinda feel like your post is an (implicit) argument in favor of the scientific method:
1. Form a hypothesis (i.e., our current best guess based on what we know, and what we think is going on).
2. Test the hypothesis (i.e., try it and see if it work :) )
3. Based on (2), go back and update our hypothesis
I think it's both reasonable and good to change one's mind over time because you've seen new data / you're in a new situation.