Thank you for chiming in. don_resqcapt19 advised that UL standards are available to read for free. After reading the standard, I see that the torque values that NEC copies on Annex I are the UL test values. In other words, you prepare the specimen to be tested using these torque values. I think I can make some sense of why UL specifies the torque based on the type of screw head being used (as compared to the conductor being clamped). When you are tightening a screw with a hex head, you can apply a lot more torque than when you are using a slotted head screwdriver. And as the comments from other electricians in this thread indicate, the typical electrician does not carry a torque wrench. So a hex head will be potentially torqued to a much higher value than a slotted head in practice. And UL tests the connector that way. But the way that NEC references this table in Annex I is very confusing. It gives the impression that the table is providing a recommended torque value in the field (which was not the intent of UL 489 where the table came from).