Apparently, the transport cycle indeed takes a couple of seconds - which is actually too long for the procedure we saw, if we assume the ship didn't actually come to a dead stop but rather did a "touch and go". Within seconds, the ship would have traveled a distance too great for the transporter to span...
However, it's not too difficult to rationalize the things we saw. Perhaps the actual transmission part of transporter use only takes a split second, even if the preceding dematerialization and the subsequent rematerialization take several seconds. This would explain both this near-warp transport and the later full-warp transports such as seen in VOY "Maneuvers".
As for the rationale for the "touch and go", we could argue that some star systems are "hostile" to warp travel, or have nasty subspace "weather". This means that one (sometimes, during the bad weather spells) cannot use warp engines within those systems, including Sol and Bajor - or at least one cannot do warp accelerations and decelerations. But perhaps just flying through is okay, and only the ramping up and down of warp fields is the part that doesn't jibe with subspace storms. It might then be a good idea to drop out of warp only briefly and not let the warp field die down. A stop of one second would mean just a stop of one second, but a stop of ten seconds might mean a tediously slow acceleration back to warp over the next ten hours...
Think of it as analogous to deploying a payload from a military transport plane. If the plane lands, the deceleration and takeoff will take several minutes. But if the plane simply flies very low, slows down to near-stall, and drops the load while still (barely) airborne, then acceleration back to cruise speed will take significantly less time.
Or think of deploying a diver from a planing boat or a jetfoil. If the boat is allowed to slow down to below planing or flying speed, a lot of time is lost. But if the boat maintains the minimum speed required for planing or flying and simply kicks out the diver, a significant amount of time is saved.
It's too bad we never saw this maneuver again. But blaming it on "subspace weather" would solve that problem: only a select few systems, at select times, would make the "touch and go" necessary. In most systems, the ship could come to a full stop, deploy the away team, and then engage maximum warp again, with little loss of time.
Timo Saloniemi