What does this BG behavior tell you?

I have prediabetes and I've been on keto for about 6 months now (recently on Carnivore and OMAD as well). I have already experienced some improvement, though not enough. Yet, there's a particular trend on my BG that does not seem to go away. Whenever I eat a meal, even if there's a good amount of carb involved (I am considering my pre-keto experiences, too), within 2 hours I get acceptable postprandial BG levels, that is less than 140 mg/dL, I suppose. I can get that even if eat a a generous amount of dehydrated grapes (pre-keto times) and have a huge spike (if it's a huge amount, then no, it won't be that nice).

What seems to me a bit weird is what takes place after those 2 hours. Then, my BG gets down, but ridiculously slowly. So if 2 hours after a 20g carb OMAD I get 110 mg/dL (peak at 149 mg/dL, 60 minutes after meal), it might take more than 12 hours (sometimes more than 16) to get back to levels before meal (about 105mg/dL or less on bad days).

Why would my body be able to do something within the first 2 hours after a meal that it just cannot do within the next 12 hours?
Wonder if there might be some clue on that kind of behavior (1st/2nd phase of insulin secretion? something else?).

My peaks two hours after meal usually take place near 60-90 minutes, I guess I've read somewhere they should appear before that (+ - 45 minutes). The only difference between keto and high carb on my curves seems to be the spike levels (lower on keto). Timing seems to be the same.

I cannot rule-out Physiological Insulin Resistance or even Dawn Phenomenon/Liver Dump, as I am regularly on mild ketosis (PIR?) and OMAD involves sleeping period (DP/LD?).

Unfortunately, I've had no glucose insulin pattern tests, not yet at least. COVID is a bit wild here where I live, so I'd like to postpone visiting doctors/labs/clinics for a while (unless it's an emergency, of course). I've also heard that whoever is on keto and on ketosis (my case) should eat "normal" amount of carbs for a few days before having glucose/insulin curves, otherwise results might be misleading. I will have to plan that.

I can confirm I have no big BG spikes. What puzzles me is that sometimes 2 hours after a low carb meal I might show BG as "low" as 100mg/dL. Yet, 3 hours after that, it might be quite the same, 99 mg/dL, 101mg/dL, and so on. Again, my system seems to be able to do something within the first 2 hours after a meal (at least BG decreases from 115 mg/dL to 100mg/dL) that it just won't do within the next hours.

But I must add that things seem to improve at night. I've noticed that after, say, 11pm, my BG starts to go down for good. That happened every day last week (even if my last meal, usually only meal, ends as late as 7pm). Around 4am, I usually wake up by myself, hours before the alarm rings, and my BG is 80 mg/dL. Last night, it was 75 mg/dL (I take no drugs for diabetes). Then it starts to raise during the morning/noon/afternoon (at 10am it's usually close to 90mg/dL), even before I eat. And when I eat, low carb food seems somewhat irrelevant, as I have no big spikes (low carb), but BG will remain at prediabetic levels or go down too slow (more so 2 hours after meal). After 11pm that is a whole other story: It goes down and might get 75mg/dL around 4am (I suspect even lower, maybe that's why I wake up for no apparent reason, maybe I'm touching hypoglycemia levels, who knows).

Does this picture make any sense? I wonder if such BG behavior theoretically points to a particular direction.