Your assumption is, as you suspect, incorrect. The atomizer (nozzle) would be at the opposite end of the tube, right where it discharges in to the furnace.
My simple test of the necked down tube was as you described. The flame (in to atmosphere, not in to a furnace) was still outside the tube, but a good portion of the atomized oil collected on the tube and ran to the ground. That most likely would not have occurred had the tube been hot enough. That was a good learning experience, however, as one does not wish to dump raw oil in to a furnace. Likewise, it won't occur if the nozzle is placed in the end of the reduced tube.
I do not. Pretty much all of this is just visualized in my head. I'll likely make a sketch of the required part when I conclude the design, to facilitate machining. I find that helps eliminate mistakes.Do you have some drawings or photos? I'm havig trouble visualizing this.
Mean time, visualize a 4" diameter tube, short in length (no more than 3/4"), just long enough to enter the existing burner assembly. It immediately reduces size via the funnel shape (likely a 45° angle per side) reducer, to the 1½" stainless pipe in which the nozzle is mounted. The length of the pipe would be in the vicinity of 4". If I incorporate a gate valve, it would be longer. The bulk of this pipe would be outside the furnace, with a port cast in the furnace in which the pipe is mounted. Same design that has been so successful in gas fired furnaces I've built. Combustion, in all cases, has been in the furnace, never in the pipe.
If I conclude that the existing blower isn't adequate, there will be no need for the reducer. That is highly likely.