Politics aside, this question also dovetails nicely with Ben's post below. I think the reason we fall into the trap of Bulverism is because, if only subconsciously, we realize that logical argument can only take us so far. We can each hold totally coherent views on a subject such as religion, even if we disagree about some priors, such as which religion is the one to follow or not follow. We can argue about who is right until we're blue in the face, but these arguments are bound to end in question-begging territory, because they are arguments about what is good (or some other abstract concept) based on differing understood meanings of the word "good". If I think that good means "whatever the Flying Spaghetti Monster decrees", and you think that good means "whatever an ideal judge would decide", then unless we can agree that the Flying Spaghetti is an ideal judge, we are at an impasse. No clever reductio or overpowering logical argument can prove one definition of good correct, because logical arguments presuppose well-defined concepts.
Obviously, this is an old problem, which is why the field of metaethics can be kind of a mess to wade into. It's easy to find problems with each definition of good, but those problems only arise because we think that this particular definition of good is problematic. This is a pretty textbook fallacy in itself, so it seems pretty clear that two people who vary on such a fundamental level will always be arguing past each other. This differing definition will totally change how they view the world, and in some sense, we might even think that they inhabit different worlds. The problem becomes, how can we still have meaningful conversations with our differing definitions of certain terms?
One solution is to go the logical empiricist route and claim that all metaphysical claims are meaningless. Unfortunately, it seems that the articulation of the verification principle is itself not a verifiable claim, so the logical empiricists have just replaced one set of prior commitments with another. We have no more reason to believe the verification principle than we do the Flying Spaghetti Monster.
Given that sort of impasse, the natural next move is to try to go in the other direction and probe at how a particular set of beliefs holds up as a worldview. If I both believe that the Flying Spaghetti Monster exists, and that pasta can never be the main constituent of a sentient being, then I believe a contradiction. Leaving aside the human mind's wonderful ability to handle cognitive dissonance and rationalize, this argument strategy usually fails about the things we most want it to succeed for. Religious tenets tend to live at the center of Quine's web (at least for those with strong opinions about religion), and so to be less subject to revision. All this argumentation will succeed in accomplishing is getting me to drop my (supposedly) rational view about pasta.
My explanation for Bulverism is that it is an expression of the above tactic. Consider what would happen if you held the following belief.
1. I hold each of my ideas because I considered it deeply and fully.
In this case, the Bulverism tactic of saying "Oh, you're only a Christian because your parents were Christian" would be an attempt to elicit some level of contradiction. Obviously, this is an easily disputed claim on it's own merits, but in this case, it doesn't constitute a fallacy, just a different level of analysis. The Bulveristic claim is not a refutation of the existence of the deity, but of the coherence of the worldview being held by the proponent of the deity. In other words, I don't want you to say "Aha, I'm a Christian because of my parents, therefore there is no God!", I want you to say "Aha, I'm a Christian because of my parents, and this means that I am not justified (in whatever way I define justification), in believing that there's a God." This suggests why Bulverism tends to be accepted for discussions about God, but not about isosceles triangles. It's a useful rhetorical device when the direct proof lies beyond the ability of two people to agree upon, but not about triangles.
This post is dragging on now, but one last point: Bulverism doesn't usually work, even in the above regard. People don't hold two-dimensional worldviews, and if they do, they have at least convinced themselves that the worldview is three dimensional. Given that, I'm out of moves. So in the comments, my question to you all is: how do we reconcile worldviews that differ based on differing definition and prior assumptions?