In reply to David Coley:
If the distance between the bolts is d, then as you say, a blown clip from anywhere between the two bolts will, before rope stretch, result in a fall of length 2d. But if the leader's waist is h feet above the lower bolt, he will end up hanging (2d-h) feet below that bolt (without rope stretch) if the clip is blown. In other words, the "clipping height" h is subtracted from the worst-case below-bolt level of 2d you'd get from falling with waist at the second bolt but no clip.
So, if you climb up to waist high to the second bolt and fall without clipping, you end up d feet below the first bolt, but the distance you can end up below the first bolt can increase to a (possibly unattainable) maximum of 2d as the blown clip is made from lower and lower positions. (All this without rope stretch.)
Rope stretch blunts the distinctions, e.g. higher clipping heights give higher stopping positions without rope stretch, but then there is more rope stretch because of higher fall factors. That said, in the range of values that are of interest, I don't think the fall factor changes all that much. The amount of variation in the fall factor in different scenarios depends on the ratio of the bolt spacing to the distance of the lower bolt from the ground. For example, if the bolt spacing is half the distance of the lower bolt to the ground, then the fall factor varies from about 0.5 to 0.7. Once the bolts are way above the ground, the difference in fall factors will be negligible.
All told, my gut feeling is that if there is something to hit, you are better off clipping high (i.e. waist close high bolt) and hope that rope stretch doesn't do you in.
This is all for sport climbing. With half ropes properly managed by the belayer, there is no penalty for blowing a clip, which actually makes me wonder whether some sport climbs with dangerous starts might be better done with half ropes.