I used to work in an industrial metal fab shop where weld testing was done, and later I worked in industrial construction where guys were welding steam pipes, including nuclear power plants, where of course weld testing was done. One of the simplest and common forms of weld testing is simply testing to destruction, putting a welded sample into a vice or press and bending it with the press or a hammer ninety or more degrees and seeing if it fails. After you watch tests like this on steel objects you learn that if someone has a basic education in welding, then they are almost always going to weld pieces together that will never fail in any normal use, let alone in destructive testing.
It is the people who have never welded much and/or do not know anything about welding steel together, that worry about it and have doubts and talk about exotic testing and sleeving etc... Also, from an engineering standpoint, almost all of the stress on a fork stem tube, is on the lower bearing, with very little on the top bearing, this is why currently there are tapered stem tubes, with much smaller bearings and diameters on the top than on the bottom, and why on motorcycles the bottom fork yoke is much, much heavier duty than the top fork yoke.
So when you are welding pieces on a bicycle fork to make it shorter or longer, you certainly do not put the weld all the way towards the bottom of the tube, you put it as high up as possible, and then the stress on the weld will be very small, not that it could not be welded at the bottom and be 100% reliable, it is just good conservative engineering to do whatever you can when you can do it.
If you are butt-welding two tubes of identical diameter, then you want a 45 degree chamfer on the tubes where they are joined so the joint is 100% weld, and not a weld that is on the surface and may not penetrate the entire thickness of the tube, not that a good welder could not choose the right temperature etc. to make a none chamfered butt weld 100% as strong as the original uncut tube. When joining tubes of different diameters, the step supplies the same sort of mechanism that the chamfer does when welding tubes of identical diameters, lots of surface area for the weld.
It is not hard to look up information on welding steam pipes, the sort of joining of tubes by butting and welding them which the lives of entire populations rely on. I have seen it done first-hand.
This is how I know that with a little education in welding and in the method used, acetylene, wire, stick, TIG etc., butt welding the steel stem of a bicycle is 100% safe and sound. If you are not educated and have no confidence then of course you should not do it. But those who know have no worries and loads of confidence. I have put thousands of miles on bicycles with welded fork tubes, and welded frame tubes, at high speeds down hills and around curves and with lots of hard braking, and have not had one worry or doubt about it, and never will.
