A couple of points of interest:
1) Even after Picard orders his forces deployed for the detection grid arrangement, the Romulans keep on considering them a "fleet", as if the odd maneuver never registered. Would the Romulans really fail to comment if the 23 ships dispersed so widely that only one or two would remain in immediate sensor range of the Romulan convoy? Or does this mean that the deployment involved only very short distances?
2) The trap that Picard creates is based on the Excalibur faking warp drive problems. Why would a failing warp drive force a starship to fall back if the grid is stationary? If you don't have warp to take you home for repairs, your reason for falling back cannot be "for repairs"! You would conduct them while doing your grid duty, unless grid duty itself calls for warp. So this might indicate that the grid is at constant movement at warp speeds, either so that it constantly changes shape in a complex warp ballet (but what would be the point?) or so that it tags alongside the Romulans who also move at warp (but everybody on the Romulan side appears to be at a standstill, and supposedly the Feds can't detect the Romulans' whereabouts for certain anyway). Or then this means that in order to be credible, the net absolutely must be capable of extending itself at warp speed at a moment's notice... But falling back still makes no sense in such a case.
3) When the Excalibur does fall back with "warp drive problems", she moves at the same speed as all the "functional" ships on the Okudagram. Then again, her drive was only in the process of failing, not gone yet, so the movement we witness could still represent at least moderate warp rather than impulse. In any case, we never get any warp streaks beyond the E-D windows, or any indication the Sutherland would be at warp, during the normal operations of the grid.
4) O'Brien says that fifteen Romulan ships are detected "spreading out along the border". Indeed, when Data fires his torpedoes, only three are witnessed close to each other. Does the expression "along the border" suggest great astrographical distances, or is O'Brien just emphasizing that the minor shuffling is not taking place across the border?
Timo Saloniemi