I'll preface this by saying tl:dr I have not answered your question, but I would point out a few reasons that make this question difficult and why I suspect you might not get a satisfactory answer:
This is more of a nitpick but "the Islamic conquests" typically refers to the period from about 630-750 when Arab/Islamic armies first burst out of Arabia and conquered the territories extending roughly from the Atlantic Ocean to the River Oxus. While there were major incursions into Afghanistan and modern Pakistan these territories were only under tenuous control (if at all) during the reign of the Arab Caliphs in the conquests. Bengal was not taken in the conquests at all.
In the absence of a formal census to calculate such things, it can be difficult to figure out when a given territory converted. I'm pretty familiar with how this has been calculated for many territories after the conquests and why as a result you might not expect to find convincing data for the territories you discussed:
In his monograph Conversion to Islam in the Medieval Period Richard Bulliett worked out a method for trying to make these sorts of estimates. Basically he relied on a few things 1. in medieval Islamdom it was common to keep "biographical dictionaries" which were these giant "Who's who" lists of society. They're slightly biased towards urban notables but they nonetheless record, at a minimum, the names and death dates of a very large group of people in a given area or field of study or whatever the subject of the work was. 2. Arab/Islamic names are patronymic, so in English they would look something like John son of Samuel son of Gerard. 3. Converts tend to either take up Islamic names or give their children Islamic names
So what he did was look at a huge cross section of names that would go something like Muhammad (d. 900) son of Muhammad (d. 870) son of Rustam (d. 820) and go "Aha! Rustam converted sometime between about 790 and 820" and when you add up enough of those names and rough estimates for conversion you can get a pretty good idea (within the limitations I've described e.g. bias towards cities/elites, no precise conversion date, whether the father or son converted, sometimes having to estimate death dates based on generational averages etc.) of when a population in total would have converted to Islam.
Even within his own work Bulliett was constrained by primary sources where biographical dictionaries were much more readily available for major centers of literacy and culture like Persia and Egypt. But his method hasn't really been improved upon. As a result for those places where primary source biographical dictionaries are readily available (e.g. Persia) we know a lot more about conversion rates than in places where they were not. (e.g. much of North Africa).
All of that is a preface to my saying that while I don't have a precise answer for you and while I'm not an expert in the regions you've asked about, my strong suspicion based on what I know of how these questions were calculated for other regions is that I suspect you are unlikely to get a better answer because 1. As far as I'm aware these regions would not have produced biographical dictionaries in large quantities 2. as far as I'm aware these regions did not have lengthy genealogical patronymic naming systems.
I hope for the sake of finding an answer that someone with a regional expertise tells me I'm wrong and tells us about the "Great Bengal censuses of 1350 and 1575" that give us a precise answer but I suspect at best you'll only find a rough estimate.
If it helps, for other regions in Islamdom based on Bulliett's work the answer seems to be "about 200 to 400 years after the region was conquered."