The question people actually want answered is how common sati was in India before anyone was counting. The returns cannot answer it, and neither can anything else that currently exists.
This is worth stating as a gap rather than as a controversy, because it is usually presented as one. There are loud positions in circulation. One holds that the practice was rare, largely confined to particular castes and regions, and inflated by colonial reformers who needed a barbarism to abolish. Another holds that it was widespread and that the colonial record understates it. Both positions are argued, and both are arguing about the interpretation of a fourteen-year series from one presidency, because there is nothing else to argue about. Neither has an independent measurement, and there is no series to compare the returns with.
What could change that is not a better reading of the returns. It is other evidence entirely. Memorial stones for immolated women survive in numbers across western and southern India and can in principle be dated and counted regionally; vernacular land, temple and family records exist for periods and places the Company never reached; and both bodies of material have been used locally without ever being assembled into a demographic picture.
Until somebody does that work, the honest statement about the frequency of sati before the nineteenth century is that it is not known, and that the tools to find out have not been picked up rather than tried and failed.