One is a sort of rambling post by Lance Mannion called The Purpose of Religion.
In it, he says
Faith, that is the belief in things unseen, is willful stupidity. Faith is believing that what your eyes and ears tell you is secondary to what your imagination allows you to wish were true instead. Faith is a belief in nonsense. To have even a little faith is to believe in nonsense and to open your mind up to all kinds of other nonsense.
That sounds harsh, and later he backs off it a little -- he seems to be trying to say that he can see the appeal, unlike some atheists. But seeing the appeal is not the same as approving.
By contrast, Pyracantha posted a nice series of short essays beginning with Science Religion Imagination Realities, part I (and continuing with part II, III, IV, and V) wherein she summarizes the conflict between those who believe in the supernatural and those who do not, and places herself somewhere outside of that, as someone who believes that the supernatural exists in our minds, and that this kind of existence, while unphysical, is still real.
She thinks that scientists will scorn this unempirical, irrational view of the world. But I think that if they do, they're hypocrites. The thing is, scientists believe in math. Math is unphysical. And yet you'll have a hard time finding a scientist who doesn't believe the person who discovered the value of pi (or rather, the means to calculate it to as many digits as you please) didn't discover something real about the universe.
I had this argument with a philosophy professor in college. I said, "Pi doesn't exist." He said, "Yes, of course it does." I said, "Show me where." He said, "It's implicit in every circle." I said, "Show me a perfect circle."
I don't remember how he answered that. It seems to me unanswerable. There are no perfect circles. Pi is an idealization, and abstraction, an idea.
The same goes for every theorem ever "discovered" by mathematicians. They are really invented. They rest explicitly on unprovable assumptions. Some of those assumptions are even acknowledged to be inconsistent with what we see about the universe around is, and yet scientists believe theorems about hypothetical spaces where triangles have more than 180 degrees to be, nevertheless, "discovered," to be in some sense, facts.
But it doesn't stop at math. If pressed, most scientists will admit that they believe in such abstractions as justice, friendship, and personal identity. All the fictions that human beings live by. You can't kick justice, you can't measure friendship, and no one is really the same person from moment to moment; nobody is consistent or predictable enough in their behavior to qualify as a real scientific phenomenon.
The metaphor that I like, I got from Buckminster Fuller. It's something like this: these abstract things are like knots tied in a rope. You can't have the knot without the rope, of course. The knot, by itself, isn't a physical thing. It's just a pattern. The physical thing is the rope. But the knot is really there, nevertheless. Another metaphor for the same thing: a mosaic. Let's say a mosaic depicting an elephant. The stones are physically there. The elephant is not, has no existence beyond the stones. It's true, there is no elephant. And yet everyone who sees it knows exactly what it is.
Buckminster Fuller said that energy was the rope, and even matter was just a knot in it. Because matter is created and destroyed (in nuclear reactions, usually) but energy, on average, is not.
You can extend these metaphors. What if I make a pattern out of knots, two and then one, two and then one? Is the pattern really there? But it only exists in the knots, which only exist in the rope. You could even imagine using the knots to make morse code, or the mosaic tiles to make letters, and spelling out the word "apple." In what sense is the apple "really there"? (It must be a little bit really there, because if I show it to people and tell them to go fetch one, they'll consistently -- repeatably! -- come back with real, physical apples.)
So that's my take. The difference between a real person and a fictional character is a level of abstraction, the difference between a knot and a pattern made out of knots.
Now believing that a God exists at some level of abstraction is not likely to satisfy a genuinely religious person, since Sherlock Holmes also exists, at some level of abstraction. The question is, is God a pattern in the universe, or a pattern in our minds? What level of abstraction? This ambiguity is sort of like a Rorschach test or an optical illusion -- is the image on the page or in your head? But maybe either way, it hardly matters. The mind is so many levels of abstraction away from matter and energy anyway, what's one more, give or take? If it exists in the mind, it's only a little further from the physical than we ourselves are. And it can still have consequences in the world, can still make people fetch apples. Of course, the only difference between a right theory and a wrong theory in science is also the level of abstraction (wrong theories exist only in your head), but if God is not a scientific theory, is more like a work of art? Does it really matter if a this still life was painted from a model or simply created in the artists' imagination? Does it matter if he got the details "wrong"?
Statements like Lance Mannion's seem to me like someone looking at a mosaic and insisting that there is no elephant, only a bunch of stones.