People also need to keep in mind that even if you have a torque wrench that's 100% accurate, the torque value you think you are applying is not what it's going to be at the threads, if any sort of thread locker, anti-seize, etc. is used.
Im not sure what the coefficient of friction of say loctite is. Say it's 0.20 (wouldn't be surprised if it was around there). You use loctite to tighten your screw to the specified 55 in-lbs. In actuality it's more like 41 in-lbs, due to the coefficient of friction of the thread locker.
Does this matter for our purposes? Not really. But I bring this up, as there's more to getting the proper torque than just a calibrated and verified torque wrench, if we are going to dive into these nuances.
Warning: we're on trajectory for landing in tl;dr territory. That will at least filter out a few of the noisier bitches and allow for a nuanced conversation among the remaining three people that actually give a shit.
I don't want to get into Threaded Fastener 101 or rewrite MIL-HDBK-60, but let's briefly state the goals of tightening a bolt:
1) Achieve the maximum joints preload that can be accomplished without unacceptable component damage. This is generally a result of applying strain to components, and
typically the male fastener represents the majority of that strain. We generally want to remove things like scope rings multiple times (mainly because most people change their optic more often than they change their underwear), and so we're not going to talk about things like torque-to-yield bolts that utilize plastic deformation.
2) Obtain sufficient friction torque to avoid unintentional loosening of the fastener. The source of this friction torque is dependant primarily on the forces generated by fastener strain; the other sources of torque (which we'll discuss more in a moment) are lost almost immediately upon cessation of the fastening process (or at least as soon as the joint is loaded in normal operation).
We measure fastening torque because it's easy to measure, not because it's the end-all-be-all method. There are reasons that one will see bolt length measured in "serious" engine builds, and engineers will develop target torque values by using test fasteners that have strain gauges inserted into the bolt shank. These techniques are not practical for most consumer applications, but they exist which proves the point.
A useful graphic to back up your point:
Underhead and thread friction torque should be self-explanatory. "Pitch torque" is what results in the stretch of the fastener, which is what we are trying to accomplish. Maximizing pitch torque is useful - recall that it's the only one that matters in operation - and we need to keep maximum total torque low enough to avoid breaking something (usually the male fastener at the point where it first enters the female fastener, which is instructive as to where maximum loads occur during tightening).
We also want to minimize "torque scatter", which is often attributed solely to the torture limiting device but generally can be applied to anything which influences the relationship between torque and strain.
Friction variation is often far more prevalent of a factor than tool variation, but lubrication can often reduce this factor which is why most "serious" work is performed with generous lube applied to the mating parts.
The type of lubricant will certainly affect frictional values - really slippery moly stuff can cut friction in half, where as Loctite claims that there is minimal reduction caused by the use of its standard products (note that there are specialized threadlockers for large fasteners that are specifically designed to also lubricate the threads). But even if we carefully take this into account, it wouldn't affect the torque we apply since the primary concern is usually fastener breakage due to excessive torsional load. The exception are situations where changes in load and/or distortion have undesirable effects - scope ring caps and action screws might be two obvious cases of this on a rifle.
Finally, note that there are several fasteners used on firearms that will come nowhere close to the ultimate strength of the joint itself, because the drive head is insufficient. If you have a 6-40 screw with a 5/64" hex socket, odds are that you'll strip the hex way before reaching the ultimate strength of the screw. Since this limits the achievable strain, use of either lubrication or threadlocker is warranted (note that each will accomplish different functions in the joint, but both may potentially reduce the risk of inadvertent loosening).