The recommendation I have for situations like this is to test it first using the test socket without any extensions connected (that is with the faceplate removed) using a temporary power supply. Now record the line stats, speed etc. (I don't know what diagnostic stats the Sky modem/router provides though). Then reconnect the extensions and perform the same action at the extension socket. If there isn't a substantial deterioration in the important stats (like sync speeds, attenuation, noise margin etc.), then it should be fine. If there is, then it might be worth looking at improving things.
However, before you perform any tests there are some things which you should really do :-
a) if there are any other extension sockets besides the one you intend to use for the modem/router, then I would strongly recommend getting a BT Interstitial faceplate filter or one of the third party VDSL filter faceplates available. Then wire those extensions through the filtered outputs.
b) if there is only one extension, then disconnect the ring wire at the master (connection 3). You will have to use a microfilter at the extension socket if you intend to use a phone there (or one of the third party filtered extension sockets).
Note that (despite what somebody has posted), running a VDSL modem off a single extension socket doesn't create a bridged tap, or at least not a significant one, as the tiny little link to the master socket is too short to make a difference. However, any other extension sockets will create bridged taps (unless filtered). Also, if plugged into the master with even one unfiltered socket attached, then that will constitute a bridged tap, which is why any master socket tests should be performed with all unfiltered extensions disconnected.
The general rule is that the VDSL modem should be the last connection on the line from the cabinet without any unfiltered branches. (Indeed, that's true for any xDSL technology).
Edited by TheEulerID (Fri 27-Feb-15 10:01:00)