Should empty columns be always extracted?

[bartekch]

Consider example

foo1 <- search_judgments(courtType = "SUPREME", limit = 1, judgmentDateFrom = "2014-01-01")
foo2 <- search_judgments(courtType = "COMMON", limit = 1, judgmentDateFrom = "2014-01-01")
extract(foo1, "division")
extract(foo2, "division")
foo3 <- c(foo1, foo2)
extract(foo3, "division")

In each case different columns are returned, because sources of judgments are different (common - supreme). Should this be unified, i. e. we always return data frame with all columns (as in third case)?