Several things here, the need for folding in old blades (not what the myths claim), The enormous differences between fusion type welds and solid state welding, how the welders handle the operation after the bond occurs, the differences between metallic bonding and molecular, and the anisotropic effects of heavy drawing and folding.
Bloomery and tamahagane type steel are slaggy messes out of the smelter, they are more like blobs of huge sintered metal particles weakly forge welded at the little surfaces which touch each other, and all the spaces between filled with silica crud. Add to this the carbon is out of control in both directions with no resemblance of even distribution. The only way to fix all of this is to squash it and press it all together while squeezing out as much crud as you can. The problem is after a few “squishes” the stuff is too thin to give you any more mass to squash, so the logical thing to do is fold it to increase the thickness for more reduction. At welding heat the carbon, over a certain level, burns out and the rest gets diffused evenly through the steel. It is impossible to get all of the slag out of it so the pockets that are trapped will be a point flaw that can give you problems. However, the drawing and folding takes these flaws and makes them directional so that they are transvers to the length of the bar stock, i.e. you would have to cross section the bar and apply force from the perpendicular direction to see the weakness caused by them. So there was a time that folding and welding did indeed make better steel, because that was the only way steel could be made, but we continually developed newer and better ways to make steel. Today we just pour it from the molten state, all but eliminating the heavy slag inclusions, and then subject to rolling operations that employ reduction rates that make folding and welding seem silly. This is why it is laughable to have a bladesmith claim he is improving a ¼” bar of steel by hammering an edge on it after it was reduced from a 3 foot thick ingot.
MIG, TIG, SMAW, O/A torch and any other modern means of welding is known as fusion welding; the metal is liquefied and allowed to flow or alloy together into a solid piece. Forge welding is a different beast altogether and is known as solid state welding. It relies on heating the metal to the point that the excited metal atoms can be brought close enough to form metallic bonds, but it is almost impossible for complete fusion and thus you will have recognizable weld zones under close examination, this is why you get some level of damascus pattern even in welding the same steel to itself. Very few industries, that I know of, utilize forge welding for critical operations, but there are cases of mechanical type welding such as explosion welding to bond metals that due to their shape or alloying would not work with fusion welding. The difference in strengths is not in the bond, of which fusion is obviously greater, but in the condition of the weld zone after all is done. Fusion welding will create HUGE grains sizes as well as heavy carbide inclusions and concentrations. Carbon does not concentrate from solid state (forge) welding, in fact the weld zone losses some carbon, and the hammering action keeps the grain size under control. So it becomes obvious why forge welding works so well in a situation like the one described by Ed with the axles. The fusion welders could have had a greater success rate if they would have cycled and heat treated the weld zones, but the forge welds did most of this wonderfully all in one shot.
Beware of discussing molecules when talking about metals, television shows get this wrong all the time, but this is because they are solely about entertaining you, not boring you with facts, so they expend VERY little energy on researching what they are saying. In steel those pesky carbides that create problems in welds can be molecular but the metal itself is metallically bonded and crystalline in nature.
The ancients didn’t come up with any super materials that somehow turn steels evolution upside down, but they did come up with some great ways to make steel with what they had. Subsequent generations built on those technologies until we got what we have today. Now we are welding with lasers and I am certain there will be even more incredible methods of bonding in the future, but they will all still work within the boundaries of physics and chemistry… the boring stuff you won’t get on the History Channel :3: