I think that speed boost Rogers and others currently use messes that up. When I checked earlier this morning, I was getting about 16 Mb/s according to Rogers and about 26 with speedtest Toronto. I have the 10 Mb/s Extreme service on Rogers. Prior to that speed boost, I'd usually have speedtest results very close to Rogers at about 10 Mb/s. That speed boost will make results more sensitive to how much data is used for the test.A speedtest from your provider is always going to be the fastest because its the shortest route.
As I mentioned earlier, different speedtest sites may give different results. This means you might want to try different sites, if you're getting significantly different results from your ISP and speedtest. As an example, there's an Oakville speedtest site that used to give very poor results, but it appears better now.Civuck, not sure what your post has to do with the OP's question?
Hugh - I think JamesK understood my post in that I was underscoring his point to the OP that different servers can provide different (often vastly) results. And sometimes geographic location isn't always an advantage.Civuck, not sure what your post has to do with the OP's question?