Suppose a part has a 5 hole pattern where the holes are for 1/4 -20 screws.For our example,the drawing might say .251 to .281 is an acceptable size for the clearance hole.The part is acceptable if a gage with 5 .250 pins on location will drop in.,but no holes plug .282.
A proud young toolmaker might make all his holes .252 and on location within .001.Thats OK,but his pride does not pay the bills.
Another man might use a .277 drill,figuring it can cut a tad oversize and still be inside .281,and he will gain positional tolerance via the hole clearance.
He can make a lot more parts in a day that gage "good".
Most AR lower assy's rattle,because most takedown/hinge pins are within .0005 of being .245.
You can build a 1911 from a bucket of mil-spec parts that will run.
Do you know,originally,the proof load was intended to beat the locking lugs into full contact?
How many mil-spec 1911's properly use the link to position the barrel up into the slide,but have the barrel underlug properly engage the slide stop to hold the barrel in lock?Both will run,which will run longer?
IMO,I do not want the slide rails fitted to .0005.It has an "ooooh,ahhh" factor,but it also has a gall factor.I might choose a .002 running clearance.
The barrel can has the od relieved behind the muzzle so the bushing is tight as the hammer falls,but there is clearance as it runs.
Its like blueprinting an engine.You make it function as designed,but intentionally put clearance where clearance is needed,and precision where precision is needed.