Think about if from an engineering perspective. The three...

Think about if from an engineering perspective. The three largest costs (OPEX) line items in REE clay processing are (and this is based on three engineering studies published for Makuutuu, Penco and Carina) mining, processing (mixing the clays and filtering them, dealing with effluents) and reagents (ammsul, amm carb, lime, acid, flocculant).

To break each down:
Mining - is 100% a cost per tonne ore not cost per tonne REE made. The grade of the ore or the leach recovery does not affect the mining cost in any way, only the strip ratio, location and mining scale do (in general terms).
Processing Cost - is probably >90% related to cost per tonne ore not cost per tonne REE made. The cost to mix and then filter the clays and move them back to the dry stack facility is based on the tonnes of ore, not the contained REE. Only the downstream MREC precipitation and filtering circuit scales with REE units produced (albeit a proportionally tiny cost)
Reagent Cost - is probably about half related to tonnes ore and tonnes REE produced. Acid, lime and flocculant costs are related to the tonnes ore treated, whilst ammsul and amm carb are related to the REE units treated.

So overall, the vast majority of OPEX is related to the tonnes of ore being treated not the REO units being produced. Therefore, the tonnage rate is the best predictor of OPEX for a project. This is supported by OPEX on a tonnage rate for the three engineering studies available for us to peruse:
1. Makuutuu (Uganda) - pH 2 heap leach - US$12.40/t ore OPEX

2. Penco (Chile) - pH 3 tank+filter leach - US$13.60/t ore OPEX
3. Carina (Brazil) - pH 4 tank+filter leach - US$13.10/t ore OPEX

The REE head grades and recoveries of the above projects vary by 200% but yet their OPEX rates on a tonnage basis are within 10% of each other - this data supports my above points.