>How are you achieving a delay more than three seconds between when the mirror is raised and shutter clicked.<
I use the Nikon MC-36 remote release (though I currently am considering purchasing a cordless remote). 1st click raises the mirror, 2nd click, several seconds later, trips the shutter.
If I was testing a lens (compared to capturing a landscape image) I would not use the remote because it is physically connected to the body and is, therefore, a potential source of camera shake, albeit an extraordinarily small one. In this situation I would press the shutter button on the body to raise the mirror, then let the camera trip the shutter itself after a 30 second delay. That, or use my PocketWizard transceivers, again for hands-free testing.
Some other points:
In order to reduce ocillation with lenses that have their own foot (i.e. tele's), I always position the foot in the Arca Swiss clamp so that the centre of gravity of the lens/body is directly over the centre of the tripod. Hence, if the head was pitched forward, the lens foot would be clamped closer to its front (camera positioned closer to the photographer). The reverse is also true.
The Nikon 14-24mm exhibits focus shift, which is to say that the plane of focus apparently moves farther from the photographer as one stops down. Since focus normally should be acquired wide open in order to minimise depth of field (DOF) and hence better judge that the plane of focus lies where intended, this behaviour can be problematic, even if increased DOF resulting from the aperture used to capture the image is, in principle, sufficient to envelop the feature originally focused on.
The issue is most likely to be of concern when the plane of focus is positioned on a feature in the composition close to the camera. This, of course, is not necessarily unusual for landscape images shot with a wide/extreme wideangle lens. The workaround is to focus one or more stops down from wide open - focus shift reduces as a lens that exhibits focus shift is stopped down. The downside to the workaround is twofold:
1. Autofocus will be acquired less dependably because the image analysed by the camera will be darker than when analysed wide open 2. DOF will be greater at an aperture smaller than wide open, which makes it more difficult for you to determine if the plane of focus lies where intended