Well, in normal industrial S/X circuits, your SX reagent (assuming it was selected correctly and is stable under the conditions of use) will last a long time. Often the diluent and phase modifiers will end up costing more than the S/X reagent when you consider that they often make up 95% of the total organic volume. So - expect to spend more on kerosene and decanol than on PC-88, for a running plant.
The cheapest effective and stable S/X reagent I know of is D2EHPA, which has been used for rare earth circuits. If your circuit can stand it, unpurified D2EHPA (which contains M2EHPA) is even cheaper. Off the top of my head I cannot tell you how it compares to PC-88 in terms of performance. In the olden days of rare earths some plants used versatic acid (a carboxylic acid) e.g. Versatic 10, which is also pretty cheap. In some places it might be cheaper to buy than D2EHPA. TBP is also relatively inexpensive and may work for rare earths. I have had bad experiences with TBP (very sensitive to silicic acid, generally requires a careful blend of phase modifiers) so I personally stear clear of it. Your experience may vary.
Just my random thoughts.