Possible Bug in the Cable Ampacity Module?

I ran a multitude of cable ampacity studies to create myself 2 tables of values with the same parameters except the thermal resistivity of the duct bank. One of them uses Concrete fill for the ductbank (Thermal resistivity = 65Rho) and the other uses Very Dry Soil/Sand (Thermal resistivity = 120Rho) for the ductbank fill. 

Given that the higher the thermal resistivity, the harder it is for heat to transfer away from the material, here's what I'd expect:

Expected:

Duct Bank Material	Soil Resistivity (C-cm/W)	Cable Ampacity	Conductor Temperature
Concrete	65	Higher than Sand’s	Lower than Sand’s
Sand	120	Lower than Concrete’s	Higher than Concrete’s

And generally this holds true for findings at higher cable sizes/runs. But at lower cable sizes/runs, I found the opposite.

Actual:

Duct Bank Material	Soil Resistivity (C-cm/W)	Cable Ampacity	Conductor Temperature
Concrete	65	Lower than Sand’s	Higher than Sand’s
Sand	120	Higher than Concrete’s	Lower than Concrete’s

You can see in the screenshot below the tables of values I've created.

I've asked SKM Helpdesk on this, for an explanation on what is going wrong, and was only told it is normal to get less ampacity with Sand backfill, and more ampacity with Concrete backfill. Followed up to get a clarification with the above table of values, haven't heard back but it's only been 2 days so I might have to be more patient. Anyone else understand what's causing this inconsistency?