Someone could always misunderstand and think it is not required to be continuous. Given that argument I would then write them up on workmanship.
I believe that most if not all of us (even the engineers and maintanance people) have all had to try and deal with the miraculous "how did they even make the original connections?" receptacle replacement.
I might not be too hard to get along with if it was just one or two and close to 6" but if the guy made them all 3 or 4" he would be fixing it and splicing an extra wire on would not be acceptable. I have never had it come up. Usually this is seen on the rough and I saw the wires being longer most of the time. If they did come up with a short one they managed to trick me Maybe John, George or Harold will have a better answer for you.
I have to agree with Gregs response. I check it on rough, and occasionally on final.
Where this arises as a situation.....is on the 'as-built' jobs that show up thru the CCO inspections. The 'as-built' basement conversions that the homeowner supposedly did. Procedure included the power be safed off, & the devices pulled to check the wiring. Then it's a red tag & a strong suggestion to contact a licensed electrical contractor to achieve code compliance. It's a real PITA.
Usually, free conductor length is not an issue with the 'sparkies', but we are all human.
As a foreman I sometimes encountered idiot apprentices who thought that our rough pulls existed so that they could earn trim scrap.
They ended up getting canned.
Short wires = more expensive work.
Yes, I've even had to re-pull conductors. ( rare )
Short taps = fumbling trim-out -- especially if apprentices.
If short taps occurred during wire pulls/ cable runs corrections were made promptly.
More generally, I never permitted any rough-in or separation to push the envelope of the NEC. Too expensive -- a route to poverty.
ONE example out of many: built up a 5,000kVA transformer room in the rough. Primary at 21.6kV Secondaries at 480Y277 and 208Y120.
Just imagine how many feeders ( 6" on primaries on down to 1.25" and 1" !! ) I allowed for an extra 1" clearance on stubs under the massive panel boards/ distribution sections.
Later, found out that the office had given us numbers in error! They were off by 1" ! I'd made it, though. My design margin had saved 850 man-hours. ( Ripping up the work encased in concrete 4 feet deep to move the pipes back into the room. )
Similar attitudes have saved my bacon WRT transformers, panel boards, the works.
If you are down to shaving inches off at j-boxes -- you are far gone. That's NOT where you make it. ( You can break it there, though. )
Same rule in Canada and code is from the entry to the box to the cut off point. Homeowners are the worst for shorting splice wire and I have definitely had a few homeowners rewire a large area for cutting the Free conductor. By practice I grab the spliced wires and if the wire nuts are beyond my closed hand it passes. Short of that I get out my tape and check.
I'd like to add my thoughts and experience. If you have a box where you have maybe three NM cables 14/2 w G, you will have 9 conductors that need to be long enough to extend outside the box three inches if the opening to the box is less than 8 inches in any direction. It is common practice to only have one of the G. wires left long to terminate on a device and the other 2 g. wires are spliced along with the long G. wire with a "Buchanan" or a wire nut.
I once had an inspection where the electrician was remodeling a basement that had been wired by a previous homeowner and the electrician cut out an old box and saved it for me. It had 2- 12/3 romex cables and an receptacle in a 1" deep box. I used it for a while in my code classes as this is what "NOT" to do. I don't know how he even was able to do it.
As for splicing a too short wire in a box- I can not find a code reference that would prohibit it albeit a bad practice. It's either splice it or relocate the box or pull new wire. I must be getting whimpy in my old age.