First of all, you can completely ignore the type of degradation you see with chlorinating liquid and bleach since that is ONLY for very concentrated solutions and varies as the square of the concentration. Even at yellow/mustard SLAM levels, the concentration of chlorine is over 50,000 times smaller than bleach and the rate of breakdown is over a billion times slower.
Losses in terms of rates are ALWAYS proportional to the concentrations of the reacting chemical species. The only time you look at absolute quantities rather than concentrations is when the reactions are so fast that the net loss is limited by the chemical being added such as adding hydrogen peroxide to dechlorinate. SLAM levels have higher chlorine bound to CYA and chlorine unbound to CYA levels so will have chlorine react faster and therefore drop faster from such reactions.
The loss of chlorine when there is no sunlight comes from the following sources:
- Chlorine oxidation of CYA.
- Chlorine oxidation of pool covers
- Chlorine oxidation of ammonia/chloramines and organics or other chemicals in the water (e.g. HEDP, Polyquat)
- Chlorine oxidation of metal and other materials
For chlorine oxidation of CYA, see
Degradation of Cyanuric Acid (CYA) where it may be hypochlorite ion that participates in such oxidation and that it would therefore occur faster at higher pH (see the fourth graph in
this post). Note that because CYA is a hypochlorous acid buffer, when the pH is higher the HOCl is largely prevented from dropping but that means that hypochlorite ion (OCl
-) rises a lot more. At normal FC/CYA levels, the CYA loss may be around 2-3 ppm per month from CYA losses I've seen in my pool over many months and repeated testing over many years which implies a chlorine loss of 5-7.5 ppm per month or 0.17 to 0.25 ppm FC per day. If one were at SLAM levels with the pH at 7.5, then this is around 10 times higher so 1.7 to 2.5 ppm FC per day. In practice, we don't see the losses be this high (which may be because my FC/CYA ratios varies from 7.5% to 15%), but if one does a SLAM at higher pH because one didn't lower the pH a lot before the SLAM than the chlorine oxidation of CYA may push one over the 1 ppm OCLT limit (remember OCLT is done overnight so normally 8-12 hours so at regular SLAM the chlorine oxidation of CYA may be 1.7*(8/24) = 0.6 to 2.5*(12/24) = 1.25). Wojtowicz in his paper claims much higher CYA losses and associated chlorine demand.
Nevertheless, there are reports of CYA loss during extended SLAM, but that also includes losses during the day that may occur from sunlight breakdown creating more hydroxyl radicals that may oxidize CYA more quickly.
In my pool, I have a pool cover and it contributes to most of my losses which with no sunlight are 0.7 ppm FC per day at regular FC/CYA levels so I figure that perhaps 0.2 of that is from oxidation of CYA and 0.5 is from oxidation of the cover. This is at 88ºF water temperature and the chlorine loss rate will be temperature dependent which is why chlorine usage drops so dramatically in colder water. Again, these are 24-hour losses while the OCLT is 1/3rd to 1/2 that since it's generally over 8-12 hours.
Also don't forget that there is a +/- 10% error on readings and while we are looking at relative readings, not absolute, I doubt people can do much better than +/- 5%. 1 ppm FC loss with such an error would be seen at an FC of only 20 ppm. So a high FC SLAM could make an OCLT difficult. This is why I think doing an OCLT at something more like 10 ppm FC is better, but we started out recommending it for a SLAM as a criteria for when to stop a SLAM. We've seen this criteria be difficult for some people to achieve, especially when they have high FC levels due to higher CYA levels. Also, one would expect to have higher CYA losses and therefore higher chlorine demand at higher CYA levels even if the hypochlorite ion concentration were constant.
The basic idea of the OCLT is sound in that it certainly detects higher than normal chlorine demand, but for "normal" chlorine losses it can be difficult to pass that test let alone measure it accurately.
The bottom line is that we don't have good data for the chlorine loss rate oxidizing CYA and the Wojtowicz numbers are much higher than normally seen.
Look at
this thread from this post on for the level of craziness demanding precise meeting of all three criteria in spite of the high FC for this half-SLAM. Do we really believe that people measuring a 10 ml water sample are doing better than +/- 5% in such measurement which would be +/- 0.5 ml? Seriously? A 5% error at 34 ppm FC would be 1.7 ppm FC which exceeds the 1 ppm OCLT limit. I'll bet that the next day this person might have the CC at 0.5 or below instead of the current 1.0, but that they won't hit 34 ppm and might be at 32 or 36 at which point people will say "you're failing the OCLT".
This is the problem with over-simplifying everything. While it's nice to have simple criteria, they can't realistically be applied across broad ranges. The actual effects are proportional so real criteria would be proportional, but of course that's more complicated not only for the extra multiplying math involved but the measurement errors that have both proportional (percentage) error AND absolute (+/- 1 drop) error.
actually I admit I'm on my phone a lot and searching for stuff takes 10X longer so I was hoping for an answer but I will continue my search nonetheless. 
This comes up often and the multiple answers JYN is alluding to are very real and somewhat confusing. We need a standard.