With some more research, titanium looks like a fascinating example.
Heavy mineral sands [0] seem to be the primary source, with total heavy minerals at ~1% of weight (all mineral components).
Of that, ilmenite [1] is the primary titanium ore (reduced to sand).
So essentially, natural primary physical reduction (of hard rock to sand) is required to meet the current market price for economic viability. There exist hard rock sources, but most would be too energy intensive to exploit, given the low concentration.
The South African Tormin operation is especially fascinating, as it has ore reduced to sand AND then washed over geologic timescales by wave action, separating out less valuable minerals and concentrating the remainder (~25% THM). [2]
Which I guess is the bulk of my point: 'at any cost', there are always more resources to mine; 'at reasonable cost', there can be sharp differentiations between different types of resources (e.g. in titanium: naturally concentrated heavy mineral sands, heavy mineral sands, hard rock).
Something being widespread in the Earth's crust, but at less than 1% concentration, doesn't help us a lot if we need civilization-scale quantities of it.
Or, if copper were distributed like that, we'd probably all use aluminum wiring.
There was at the time a severe shortage of usable titanium refined metal. Refining Titanium is much more difficult than aluminum.
The soviets had over invested in the ability to produce it, so it was more economic to get it from them than try to produce the capacity here.