Oh, not seeing an arguement here.
I went to 33 on low carb/keto. Low GI I did try out right at the beginning, but it still messed my sugars up before I realised it was all carbs, fast and slow, that did me in. No idea whether low GI would've made a difference if I'd tried it sooner though...
I have recently been recovering my bgl by using low carb after release from hospital which had messed my bgl control summat chronic. I was seeing 5's and 4's, except for the morning, where I seemed to be experiencing Dawn Phenomenon because my fasting levels were significantly and consistently higher than my bedtime readings. My late night readings would be (per ex) 5.1, and my morning one would be around 8 or 9, then my midday level would be back to 5 or 6.
I finlly worked out what was happening, Because I am using a high dose of Gliclazide, I was being proactive at night by having a low GI snack at bedtime to make sure i did not hypo in my sleep, This is the cause of the morning rise.
I proved this by stopping the snack, and seeing that my morning fbgl came in almost identical to the bedtime reading, i,e, unchanged overnight. So it seems that the Glic has run out of steam by the time I went to bed. This gives an active period of around 4 or 5 hours for the med to be neutralised, I also proved it to be due to the Low GI snack by using a bedtime reading that was significantly higher than usual (due to bad food choice), and then eating a large dose of milk chocolate at bedtime. Lo and behold, my fbg the next morning was 5.1 again, following a bedtime reading in excess of 10 mmol/l. So this gives me evidence that sugar triggers my Stage 1 insulin response, and it works fine at night, It is NOT triggered by a low GI snack. The low GI carbs will push the bgl up overnight and not be recognised by my body as being unwanted so will give a raised bgl reading in the morning. My breakfast and morning meds will eventually deal with the low GI snack by midday, but I have a long acting glucose bump following low GI, as it creeps in under my radar,
In hospital I regularly dropped from a high bedtime reading to a low 5 ish fbgl in the morning, and I put this down to having a bedtime cocoa drink, with sweetners to help me sleep, This seems to have been enough to trigger a Stage 1 response. I am not sure if it is due to the Cocoa, or the milk, or the sweetners, since artificial sweetners are suspected of triggering Stage 1 responses on their own.
So maybe a way to deal with DP is to eat a sugary snack at night, which seems counterproductive but may restore the fasting levels. Its a cheat since all it does is push the sugar carbs into storage to lower the bgl, Alternativly make sure you do not eat low GI at bedtime.
This post does not address the OP regarding reversal etc, but I regard that argument to simply be semantics, and I do not feel inclined to join in that debate. However, the experiment to trigger Stage 1 response does go partway to demonstrating that my IR has been reduced, and that I may not be so far away from Diabetes Nirvana after all, I produce my own insulin and it does what it says on the tin!!!!!! Now that I know that the med only lasts a short while, I am also in a better place to reduce it ad absurdam and I do not need to fear nightime hypo's either.