I'm a researcher at Curtin University working on the Desert Fireball Network. The DFN is the largest fireball observation network in the world, and our primary goal is to recover meteorites with orbital information attached.
The traditional rule of thumb for meteorite-dropping events is a final luminous height below 35 km and a final luminous velocity below 10 km s-1
So when I refer to the luminous part of the trajectory, I am referring to the portion in which ablation is occurring and optical light is emitted. As you said, this portion typically lasts a few seconds depending on the initial size, strength, speed, and slope of the meteoroid. After this, if the object still has mass left, the speed continuously decreases. We call this portion of the trajectory the "dark-flight" (as opposed to the "bright-flight" which is the luminous part) because we can longer observe the meteoroid.
During the dark flight, the rock goes from <10 km s-1 to tens to a few hundred m s-1 when it finally impacts the ground as a meteorite. This part of the trajectory can be greatly affected by wind as seen below from the dark flight modeling done for the recovered Dingle Dell meteorite:

So, in total, from first becoming a meteor to impacting the surface, you should expect about tens of seconds to minutes to pass. Of course, this is dependent on things I have mentioned like mass, speed, and slope notably.