I guess it’s purely academic, but does it somehow make me less of a 3d printer guy if I have to scale my parts 106% to be dimensionally correct?
106% ?? Depends on your material.
I do 102% when I need dimensionally accurate prints in ABS.
I didn’t vote - the right answer is just “No”
Nah, that’s pretty much true for all of us running some sort of reprap or derivative. Holes are too small by 0.4 - 0.5mm. But the Z axis is accurate. So I think using the XY dimensional compensation in slicer will be better as you can measure your consistent error and pass that along for all future prints.
I guess it depends on why you’re having to scale then. I do it to compensate for material shrinkage during cooling. It applies in all directions.
Thermal correction makes sense…I’ve adjusted flow down to 95-97% for things like print in place bearings, but it won’t nicely fill the top layer of normal prints.
It is well known that plastics will experience shrink after being injection molded. This it’s just your way of compensating for plastic shrinkage in fdm!
And if you’re printing for lost-pla casting of aluminum, make it 112%.
You could adjust your machine description to hide the scaling. But either way is OK in my book.
@paul_wallich how/where do you do that? In slic3r, preferably. I always print abs on one of my printers and only remember to scale about 80% of the time. 
I’m guilty of doing this too. BUT…what most people don’t realize is that dimensional accuracy can often times be corrected by simply increasing the quality of the STL file when you export it. But yes, plastic does shrink and you can’t always help that even if you have the highest quality STL file possible.
I just have a tol variable in CAD (inventor) that I set to the diameter of the nozzle and add to the diameter of all holes. So that works out to 0.4mm extra clearance for a tight fitting hole and an extra 0.1 for a loose fitting hole.
I’m just now starting to experiment with the XY factor as I’m using different nozzle sizes for different parts and across different printers.
But when I printed my iPhone 6 case, 1.006 scale factor was the magic number.
1.006, or 1.06?
I also use 1.06 as my scale to get a more correct object. This is particularly annoying since I am making boxes and slots for objects where I know their size ( with a caliper)
I don’t think that’s reflective of normal FFF materials. For example, I recently scaled a part up by 1.005 (0.5% larger) to compensate for ABS shrink. A material with 2% shrink is unbuildable without an enclosure. I would get a dial test indicator to measure the actual machine motion, count the pulley teeth, etc. It’s like you’re using a drive pulley that is one tooth smaller than the firmware is expecting.
I scale 1.006 for ABS, 1.003 for PET, and 1.002 for PLA. Shrink factors for our plastics aren’t on the order of 6%, more like fractions of a percent. Once I use the single wall cube to properly tune my extrusion multiplier for the spool, and the shrink factors above tested on calibration object, I usually come in within 0.1-0.2 mm on dimensions ID and OD.
No, you’ve chosen the simplest path to the best result. It might cause issues down the road, but for now it’s an excellent and efficient solution.
I never encounter that much shrinkage.
Yeah guys 6% sounds awful and way too high.
My decimals were correct, 1.006 scale factor.
For me it’s more like a factor of 1.01, but only on X/Y. And I know that I could probably get that compensated for by more calibration but honestly, I’m being pragmatic here - if it works reliably by just adding a percent and otherwise everything is ok, I’d rather not touch a running system 
I add 0.5mm extra clerance in the model design, that makes it fit perfectly.
Also, nopheads hole sizing math is interesting here.