Yeah, a ti reusable container would make things far better, as far as canister use goes. The cans are heavy and the waste is quite high given the contents, not quite doubling the weight of fuel. The gadgetry in the valves is also wasted labour, well at least for robotics. Far simpler to use a refillable propane bottle of around 240g capacity and eliminate the need for butane and isobutane cans. (But, this would also eliminate the need for getting rid of the butane - generally a byproduct from the petrolium refineries, often just burned off with excess methane, et al.)
All this waste is my primary objection to canister stoves. The cans need not be ti, and could be aluminum, as Roger pictured in the article. And, useing propane prevents the winter "freeze up" that is common with toppers. Titanium is actually a heavier material than aluminum, is far more expensive, and, is more difficult to work with (spinning/shaping.) It IS a bit stronger, though.
As far as stove systems go, there are three primary components to all stove systems. 1) some sort of heat source. 2) Some sort of heat target or pot. Some sort of energy conduction/radiation to get the heat from 1) to 2) with resonable efficiency. Optimizing any one component is sort of a worthless ocupation. All three need to be optimized together, ie, balanced together. This is, perhaps, why I find the 300t so interesing. It utilizes the pot as part of the stove. That is, the heat target is integrated into the heat source. It lacks a heat exchanger, IR absorber, wind screen, but there are potential solutions out there.