Eh. Today's Gini coefficient is 0.47, the estimated one back then is 0.44. Given the uncertainty of the data, that's pretty much a wash. Also, they're not factoring in the slavery aspects of slavery into this. And of course, it's not that surprising - the US was very egalitarian back then, something which isn't news.
The data are not very comparable, that's for sure.
I would add (mostly for pedantic purposes) that Gini is a somewhat arbitrary measure of inequality. It is about as good or as bad as any other measure that would satisfy Lorenz criterion. There is really nothing particularly intuitive about it, except that it is a) a measure of something about incomes; b) anonymous (just counts people and their incomes, ignores their names); c) scale-invariant (doesn't depend on absolute levels of income, just on its distribution); and, most importantly, d) it satisfies the transfer principle: if you transfer the money from the poor to the rich it grows. All those are desirable properties of an inequality measure, but there are many measures other than Gini that satisfy it as well. Hence, even if the data were perfect, minor differences in Gini are not very likely to mean much, unless you can show that other possible measures (such as enthropy) show the same.
