Firas : yes
David : TmBo is probably talking about the “outage” due to the HO procedure, or perhaps the duration between the HO Request and the HO Command. His post is a little ambiguous though, I agree.
The outage is not linked to the average window, it is due to the time the MS takes to move from the old TCH to the new TCH.
The duration between the request and the command depends on where is lcoated the target cell : same BSC, different BSC, different MSC, ..