I have been waiting for a thread to appear for which I could respond with some degree of technical authority. This may be my one and only hope, so please be kind.
When I first started my apprenticeship, the pipe dope versus Teflon tape debate was an important part of my training as an outside machinist.
Short answer: pipe dope is better at stopping leaks
and does a swell job when properly applied.
Long answer: Teflon tape has a few purposes but why it works is as important to understand as is how to apply it correctly. Most do not use it right or so said my mentors. What it does is reduce friction. That allows threads to seat further into the wedge. It also stops galling (meaning metal on metal seizing) so that assemblies can be taken apart at some future point.
To proper apply Teflon tape, it must be wound such that the wrapping tightens as it is inserted into the female threaded part (nut, tapped hole, cap, etc.) One and a half wraps is all that it needed, (manufacture spec) because using any more than that becomes packing. Excess packing reduces the amount of room that the threaded parts need to seal and actually lessons the ability of the joint to seal (more on this latter).
Rather than blindly accepting the opinions of our mentors, we noobie technicians actually conducted experiments. Lots of experiments! Teflon tape does work on machine screws (straight threads) as well as tapered threaded assemblies. The reason this is true is because as the assemblies are tightened, the interface between the treads of the fasteners actually deflect.
The threads themselves are bendable (if only slightly). What gives them their gripping force is Mr. Inclined plane. The threads are literally wedging a tapered shaft into a valley. There is an air gap between the peak of the male thread and the valley of the female trench. But it is the side to side surface tension that seals
the two surfaces; the side to side contact of the threads (the force of surface tension) is what makes the threads hold fast.
Adding pipe dope also decreases surface tension so it allows the threads to be tightened further. Exactly where in the system of thread to thread contact that the most friction is created is not easy to judge. So we used bluing agents to study the interfaces after partial and full assembling.
The friction could be in the middle, or towards either end of the fastener but also on the tips, center, or nearer to the roots of the threads themselves. Adding a lubricant like Teflon tape, or pipe dope, eases the tension of deflection in the joint and smooths out the distribution of tension so that more side to side contact occurs (meaning over a greater surface area).
Which is better is somewhat dependent upon the purpose. But for us, we used both strait threads and tapered threads as well as capped tubes and pipes of various sizes in our test samples. These were put under water pressure and allowed to fail. We tested properly torqued assemblies, over torqued, under torqued, and all of the above with combinations of Teflon tape, pipe dope, and both pipe dope and tape. We tested no wrap, over wrapped, properly wrapped, and under wrapped, as well as thin, correct, and too thick of pipe dope application.
Then we intentionally damaged the threads in various ways and repeated the tests. It took a full week to do all of the tests but the longer term examples stayed in place for months and even years. We had a pipe shop with calibrated pressure instruments to use to monitor
the assemblies. Some of the tests from previous year's classes
had stayed under pressure for 10 years; they let us take them apart just to teach us the differences.
To me, pipe dope is the clear champion against leaks
. I know this because I filed a groove into the entire thread length of a tapered assembly and tried my best to seal it with torque, (epic fail), then Teflon tape, (serious leakage) and then pipe dope (winner). I even made runs with tape and pipe dope but that too leaked rather quickly. To my surprise, the pipe dope only samples held for weeks even though there was a complete channel filed into the thread of the male part of the fitting.
By the way, over torqued consistently failed sooner than proper torqued joints but adding tape, or any other lubricant such as grease, does change the amount of torque that needs to be applied or else thread damage can (and did) occur.
You are welcome to repeat your own analysis, but the results will probably not vary. At least they didn't for my class, or the ones before mine.