My mindset, incidentally, is snug with the impact but torque with a torque wrench. Lug nuts/studs are a torque critical fastener in that the torque spec is usually very close to the yield point of the stud. You can't tell if a stud is stretching and failing if you aren't torquing it properly, and I've seen a lot of cheap replacement hubs/studs where the supplied studs would stretch and fail before even coming up to the specified torque.
On the other end of the improper torque spectrum, there are significant shear loads on the wheel/hub interface (under braking, mostly, but also under acceleration) that can work the stud back and forth as the wheel slips fore and aft. Any time you see a stud that can free-spin loosely in the hub because the knurling is damaged, that is from being run undertorqued. Especially in aluminum hubs! This of course assumes that it is only lightly undertorqued, and not nuts backing off undertorqued Best case you have to replace the hub, worst case you shear the studs off from them constantly being bent back and forth.
This is why high powered drag cars don't use tapered lugs at all, they use flat lug nuts and drive shanks on the studs to locate the wheels. It's not perfect but it's a hell of a lot better.
Vigo
PowerDork
7/24/16 8:39 p.m.
TLDR but after many years as a professional tech and a still-current ASE Master Tech + L1 and vocational automotive instructor, yes you can do an adequately good job to tightening all sorts of things with an impact, after accruing a lot of experience. I think it would be difficult to get that experience without doing it for a living unless you are a very prolific enthusiast. Before cordless impacts hit it big i would sometimes use my 1/2" impact with a 1/2>3/8 adapter to snug up worm gear clamps. With enough good judgment and an impact with fine trigger control, yes you can do it.
However, at work I always instruct students to use a torque wrench for everything. Many of them ask why it is that people in the field commonly don't use torque wrenches for most things. I tell that over time you can develop a 'calibrated torque arm' but that you will NEVER learn what XYZ torque 'feels' like unless you spend your formative years clicking a torque wrench and developing an internal spreadsheet of 'what torque does this fastener likely need' based on diameter/length of engagement/thread pitch/material it's threading into/bolt hardness etc etc etc. Another thing you have to develop a good judgment of over time is how the length of the tool handle distorts your perception of torque.
Another thing i teach that i feel is very important to realize is that when you 'snug' something by hand or with an impact and then follow up with a torque wrench, you need to see that fastener move before the wrench clicks. If the wrench clicks without the fastener moving, you may be clicking at 90 ft lb on something that is torqued to 95 or 905 or 9005 ft lb. If you don't see it move, back it off and then torque it again and try to make sure you didn't do any damage that now needs further work. That right there is a learning moment if you stop and think about it. Stopping and thinking is one of the hardest things to force yourself to do, especially when you are racing the clock as a working flag-rate tech.