Every tank is going to be different, but we can make a general outline of the possibilities.
The water coming out of the reactor likely contains more bacteria and bacterial byproducts than the tank water. That's why the reactor is there, fundamentally. Those bacteria and byproducts can be a food item for a number of organisms in the tank. So letting all or some of that food reach the tank might be useful. If the amount of food going into the tank is low enough for the tank to consume it, then you might see more growth in the system.
On the other hand, if the food reaches the tank, and it just decays back into nitrate and phosphate, then the utility of the reactor might be, well, limited.
Skimmers are not all that efficient, from what data we have, so I'd expect them to require multiple passes to remove the food in question, so feeding the output of the reactor to the skimmer might be just fine on both counts. It could leave some food in the water column, and remove enough so that the reactor is useful.
So what would I do? I'd likely just set up the reactor and let it run for a while, to see how it does. If the parameters aren't dropping or are dropping too slowly, getting more of the output into the skimmer might be worth some effort. If the tank was having serious problems, though, I'd route the output to the skimmer from the start, and work on any possible tuning later.