The two tables just have different resolution at different locations, that's why they look so different. But the span is near identical, differences are likely just different teams doing the calibrations and using different levels of precision. The little thermistors inside are the same so you could exchange one table axis+data for the other and nothing would change, not that'd be detectable anyway.
For the longest time GM used the same 100k NTC thermistor element in every single sensor that measured any kind of temperature - ambient, interior, coolant, intake air, everything. Plastic body IAT or metal body, open element or closed, discrete IAT or integrated into the MAF, all the same electrically. Now, they may have fiddled calibration values here and there to tweak things, but the sensors (the '100k' ones) have the same guts in them.
The later MAFs with the axis that goes from 50 to 43k, I believe those are using a generic '50k' thermistor element. So now there are two...
edit: and... these are likely generic commodity parts, with something like +/-10% tolerance. Swapping one sensor for a different, identical sensor could have more difference between them than what's in two different tune files both using the '100k' thermistors.