[A_D_E_P_T]

> the oxides of aluminum, magnesium, and nickel were not in use as paint pigments

Aluminum oxides were used as a pigment, predominantly in blue (cobalt aluminum oxide) but also in white.

In any case, the dominant white dyes of the Early Modern period -- and prior periods -- were lead based. The presence of TiO2-based pigments is actually one good way to identify a modern forgery.

> the particularly relevant issue here, as i understand it, is that titanium has a stable carbide

This turned out to be solvable via calciothermic or magnesiothermic reduction -- which is now effectively the go-to method for just about everything that can't be reduced with carbon. All titanium dioxide reduction processes demand quite a lot of energy, though; more than aluminum and far more than iron.

[kragen]

people don't make titanium via calciothermic or magnesiothermic reduction of the ore, i don't know why

the magnesiothermic reduction is the actual reduction step of the kroll process, though