Okay, so, here is my scenario:
Imagine a group of miners who are mining a coin are responsible for a large portion of the hash rate on the network (let's say not quite 50%). Looking at the parameters of a coin, one can easily deduce when difficulty re-targeting will occur.
So let's say the group of miners stop mining the coin for the duration of a difficulty re-targeting window, causing a sharp decline in the hash rate, and thus very likely the rate at which blocks are being found, ultimately causing the difficulty to drop. They then all band together and start mining the coin again, with the significantly reduced difficulty and are able to mine blocks much faster for the duration of the difficulty re-targeting window.
Is there any known cases of this occurring, and what preventative measures could be taken to avoid this? I realize this is akin to a 51% attack, with a few key differences. Any thoughts/comments would be appreciated!