This is one of my pet peeves, a problem in the Nikon flash system. All the lens can measure is focus rotation, supposedly corresponding to focus distance. It's not too bad in prime lenses, but the cheaper zooms are terrible. They focus differently at different zooms, so it's a crapshoot, mostly a far wrong distance is reported.
The bigger problem is the stupid Nikon flash system uses it anyway (you'd think their engineers would know it's wrong). Only affects TTL BL direct flash, not for real TTL mode or bounce, etc. But direct flash typically sees a dark background (inverse square law), and a dark background will generally overexpose a shot, trying to brighten it.
So the bright idea is that if the metered flash reads too strong for the guide number distance (the distance indicated by the lens), the flash is reduced accordingly. Which seems a good thing if the lens distance is accurate, but usually it is not, so we get dark flash shots (TTL BL direct flash). No harm if the lens distance is greater than true (it is ignored), but it screws up the shot (TTL BL direct flash) if the reported distance is shorter. And it often is, esp at wide angle zooms. This is a routine default case, TTL BL direct flash, wide angle (but longer zooms are not immune).
So, it is a genuine flash problem (TTL BL direct flash) that we have to overcome. The flash head tilt switch triggers this issue.
A page at
Nikon TTL BL flash - D-lens distance data accuracy is about this.
Also non-Nikon flashes seem not to suffer this problem (
Non-Nikon brand flashes solve the TTL BL Zoom Problems ) ... no head tilt switch.