With all respect, I haven't noticed strength dropping considerably, unless I did something stupid and worked very close to failure recently enough. Then yes, my 5RM could drop to 4RM or even 3RM between bouts. All in all, without hitting the failure spot and assuming sleep/food is fine, my 5RM from previous w/o is at least my 5RM on subsequent w/o. This can be explained by my relatively low training age, where progress comes easier.
So when you say "If you normally work with 5RM loads", you mean you that you're used to working with sets of 5 reps at an RPE (rate of perceived effort/exertion) of 10, whatever that load might happen to be on the day.
The fact is though that you can't always easily gauge accumulated fatigue, so even though you feel that your recovery is good and that your sleep/food etc. is all spot on, you still might be more fatigued than you realise compared to your previous session.
One way for you to test for this is to do a set with the load from your previous workout as your first work set; if you manage 5 reps with that, increase the load for your remaining work sets and see how you get on. Sometimes the same load will 'feel' much heavier than it did on the previous session due to neural fatigue factors. This test will help you to prime your CNS and psychologically you know you lifted that same load for 5 reps the previous session.
If you find you can't manage 5 reps with the load from the previous session then you know you have some kind of accumulated fatigue taking its toll. In that case, stick with that same load and bang out your work sets with that. Try again the next session. If the first set of 5 goes up without too much of a problem, add a bit of weight to the bar for your remaining sets and see how you get on.
I'm also not sure what you mean by: "Working true 2RM or even 4RM might take some learning"?
Meaning: according to
this calculator, having done 77 kg x 5 I can do 80 kg x 4, not 2 like I did twice, and 3 reps a week later.
I can explain that to some degree by accumulated fatigue and CNS stress, but more so by lack of considerable amount of training with doubles and triples. On the same note, if a person got used to benching 100 x 10 with great form almost to failure, doesn't necessarily mean he can bench 117 x 5 without first getting used to training at lower reps. AFAIK this is called SAID, or specific adaptations to implied demands. Body learns best to perform a specific task we're putting it through. Well, this is a contrived example, as a person training on HST would do both 10's and 5's regularly, but for a person not following HST and working mostly in the 8-10 reps range this is more so.
I question your use of the SAID principle with your benching example. Your muscles don't care about reps. They care about how much work you ask them to do and for how long (TUT). Try doing 5 slow reps with the same load you can just manage for 5 fast reps. It's not happening.
I suppose a considerable amount of training with doubles and triples might make you better at lifting max doubles or triples than if you mainly trained with max sets of 5, but not necessarily so. Once you are experienced enough to maintain good form for a max set of 5, it should carry over well to 3RM, 2RM and even 1RM attempts.
As a case in point, many powerlifters do not test their 1RM in training because of the greater chance of injury and the high level of fatigue induced from such attempts. It is much more likely that someone would train with a load around their 5RM for triples so that their form would remain as consistent as possible.
What might be happening in your case (if we rule out fatigue etc.) is that your form is altering a little with the heavier load and you might be consciously (or subconsciously) moving more slowly. It is inevitable that you will move more slowly as the load on the bar increases but you might be exaggerating this a little through a fear of getting hurt. The trouble with any true RM attempt is that your form will usually start to break down for the last rep or two, depending on how many reps you are doing. A less efficient bar path takes more energy to complete a rep so further reps are harder.
NB. Not all rep/load predictors use the same numbers.
Reps 1 2 3 4 5
Brzycki 100 95 90 88 86
Beachle 100 95 93 90 87
Dos Remedios 100 92 90 87 85
As %ages of 1RM