One reason for adding a delay to the main speaker output is to compensate for the difference in length of the acoustic paths from the sub to the listener and from the main speakers to the listener. If the sub is behind (further away than) the main speakers, delaying the main speakers compensates for the fact that it takes longer for the sound to arrive from the sub.
From your description I'm not sure whether the 2m is actually the difference in path length as @chrisc has suggested - rather it sounds to me as though some trigonometry is needed. Would you be able to post a diagram of your set-up with measurements? - it should show the position of both speakers, the sub and the listener.
As well as the difference in the acoustic path length, the delay can also be used to compensate for signal path delays (e.g. due to DSP) in the sub and to match the phase of the sub with the main speakers around the cross-over frequency. I did quite a bit of reading around this topic when I ran a sub with my previous set of speakers and concluded that there was no "on paper" way to work out the correct delay -- it needs some experimentation and listening, although setting the delay to compensate for differences in the acoustic path length might be a reasonable starting point.
I'll try to dig out the articles I found when I was looking into this, and will post links here if I can unearth them.
From your description I'm not sure whether the 2m is actually the difference in path length as @chrisc has suggested - rather it sounds to me as though some trigonometry is needed. Would you be able to post a diagram of your set-up with measurements? - it should show the position of both speakers, the sub and the listener.
As well as the difference in the acoustic path length, the delay can also be used to compensate for signal path delays (e.g. due to DSP) in the sub and to match the phase of the sub with the main speakers around the cross-over frequency. I did quite a bit of reading around this topic when I ran a sub with my previous set of speakers and concluded that there was no "on paper" way to work out the correct delay -- it needs some experimentation and listening, although setting the delay to compensate for differences in the acoustic path length might be a reasonable starting point.
I'll try to dig out the articles I found when I was looking into this, and will post links here if I can unearth them.
Roon (Mac Mini), Wilson Benesch Full Circle, Expert 1000 Pro CI, Kaiser Chiara
Warwickshire, UK
