Darn, I was hoping for a different number from the Refugium water. Okay that points the resolution in another direction. Let's look at tweaking something that will influence the entire system.
Good changes so far. I always run the RDP lights overlapping the main lights by a few hours. 24/7 lighting is not necessary here and because the algae is not growing well anyway, I don't believe it will help at this point.
I was being a bit dramatic regarding my no vacuuming comments. I've had too many occasions talking to hobbyists telling me about their really dirty tank and how they vacuumed it often to keep it clean. Of course this only compounds the problem, requiring more and more vacuuming.
Yes, from my experience I would definitely remove the dieing algae. It's just putting unwanted stuff back into the water. I suspect that algae harvesting has not been part of the regimen in this system, right? Because it was not growing. And yet the test shows higher levels of Nitrate. This is the red flag and probably the reason for this thread.
Have you considered using some Chaetomorpha or other Macroalgae along with the Caulerpa? Different Alga can process N at different rates.
If the PO4 test shows nil, that could be a reason for dieing algae and higher N. The lack of Phosphate leaves the algae starved and it cannot grow, thus N stays in the system longer before being processed by the other parts of the biofiltration. The addition of foods containing more PO4 or a chemical form of PO4 could do the job. There are fertilizers meant for Macroalgae. The "f4" fertilizer used to grow phyto could also be used. I've even used Miracle Grow fertilizer. Algae needs iron to grow.
Coil denitrators work extremely slow. I wouldn't waste my time on this one.
Clean sand will start to populate immediately and it's effect will show as early as one week. Oolitic sand is especially useful since it requires less sand to do the job. If this were my tank, I would do two things first:
1. Add Oolitic sand to the tank, placed under the existing sand. (Make several depressions down to the glass, slowly pour Oolitic into the depression and cover it over with the old sand.) Since most Nitrates are processed in the deepest area of the sandbed (also in the deepest part of LR) this action goes straight to the need, to reduce Nitrates.
2. Add another type of Macroalgae. This will help the Refugium. Caulerpa is so prone to go sexual anyway, that the additional variety of algae not only seems to inhibit the sexual dissolving of Caulerpa but it adds to the N processing ability of the Refugium in general.
3. I would consider a tiny amount of PO4 and Fe rich fertilizer and watch for an immediate spurt in growth.
Adding a sandbed to the Refugium is optional at this point and if it were me I would hold off just to see what difference items 1, 2 and 3 make.