@cernel Ok, I found what appears to be a fairly reliable reference (Princeton University, New Jersey):
https://irs.princeton.edu/sites/irs/files/Rebasing Maddison_May_2017.pdf
An important implication of using different relative price levels is that the poverty level may change. With the 1990 price levels, the subsistence level income was estimated at between 350 and 400 international dollars per year (Maddison, 2003). The poverty line was equal to around $ 1 per day, and was based on the first international poverty line which was set at $1.01 per day using 1985 PPP’s, which was later updated to $ 1.08 per day using the 1993 PPP’s (Ravallion, Datt and van de Walle, 1991; Chen and Ravallion, 2001). This made the interpretation of historical income series very intuitive. By using other relative prices, this subsistence level of income changes. The price level (in US dollars, the standard used in these calculations) increased by 59% between 1990 and 2011, bringing the poverty line to 636 dollars of 2011. Moreover, The World Bank slightly raised the absolute poverty line to 1,90 US dollars a day or 694 dollars per year, expressed in 2011 prices.
694 United Statesian dollars looks about what I had in mind; so, in this case, the approach that I suggest you using is to calculate what I would call the "GDP Above Subsistence" (practically, the "usable" GDP), by the equation (assuming using 2011US$):
GDPAS = (cgdppc - 694$) ⋅ pop
Still, India is going to be overvalued, since it was a country with a lot of people and that was somewhat decently productive (surely in a better shape than China), but that, at the end, contributed relatively little to the war effort of the British Empire (and had some major famines, as well), at least economically.
