Thanks Grant,

Thanks for the help.

On thing that jumps out on the waits is LCK_M_SCH_S with with a average wait of 119 secs. I have noticed that the Redgate tool takes a lot of these waits though, without any performance issues so I might be barking up the wrong tree.
The highest is the CXPACKET waits. Even thought at one point it it might be a parallelism bug, but probably unlikely...No other waits really to take note of.
Might be some outside resource contention, but looking at the redgate tools everything looks fine (No CPU, IO, memory contention). In fact, these all goes down during the slow running query?