Lathe beds & headstocks made of fabricated steel are nothing new. The method was used on some very heavy duty high production lathes by LeTorneau. The problem with any weldment is post weld stress. The frame of an H-D is lighter material, and an automated, sequenced welding procedure is used to minimize or control this post weld stress. On something like a lathe bed, heavier steel plate would need to be used. This requires more weld, to get full penetration welds. A process such as GFCAW (gas shielded flux cored wire) would likely be the choice. Once the bed or headstock was fabricated, it would then need to go into a stress relieving furnace. After stress relieving, probably a shot blasting to get it cleaned for rough machining. After rough machining to within probably 0.030" of finished dimensions, back into the furnace for a final stress relieving.
LeBlond, on their later lathes, got the best of both worlds. They built the lathe beds out of good iron castings, and used hardened steel replaceable bedways. These bedways were precision ground. We have a 25" x 96" NK series LeBlond (wide bed, heavy duty) and a later 15" "square head" Regal at the powerplant machine shop. Both have the replaceable bedways.
The other property which cast iron exhibits is great dimensional stability and great vibration dampening. It is an ideal material for lathe beds, headstocks, and other parts. One pour and a bed is cast, no cutting multiple parts, jigging them for welding, welding.... With today's "burnout molding" (using rigid foam patterns), a lathe bed or headstock could be cast without needing to make a bunch of dry sand cores. I believe the Chinese are casting their machine tool beds and milling machine mainframes for Bridgeport knockoffs using this process.
With any automated welding process, it comes down to accessability for robotic welding. Something like a lathe bed, with multiple "girths" (cross bracing) would be a bit trickier to get into for robotic welding. On the other hand, foundries are well automated. My wife bought a bacon frying pan that was on sale, mainly because it was made in USA. It is a "Lodge" brand pan, and they have a website. They have some youtubes of their foundry, and the degree of automation is amazing. I am sure that for cylinders and heads, H-D uses an automated casting line.
What I have noticed in the Chinese casting used on some of their machine tools is that they lightened the castings. Instead of casting heavier sections, they tend to go lighter and thinner. In addition, they use a lot of thin ribbing. The surface on the unfinished sides of these parts has me thinking they are using the burnout molding processes for some of the parts. I think the Chinese are also using an alloy of iron and steel ("semi steel") with a lot more scrap steel in the melts. This gets the tensile strength up, and lets them get away with thinner castings.
The old line US Machine Tool builders poured iron from cupola furnaces (melted with coke), usually reclaimed from scrap (such as old engine blocks, old home heating boiler sections, busted machine parts...). If they used any scrap steel, it was a much lighter percentage. The result was an iron that was much more dimensionally stable and had the better dampening properties. Then came the matter of "seasoning" the iron castings. Most machine tool builders got their castings in a run from the foundry, and stored them outdoors. Some manufacturers deliberately let the castings sit outdoors in the weather for a couple of years. The belief was that this "relaxed" the iron and made it more dimensionally stable.
Automobile engine manufacturers did this same thing. The business of "just in time" manufacturing was unheard of and could not be done if "seasoned iron castings" were going to be used. The machine tools builders probably knew their business was cyclical, and might be "feast or famine". As a result, they probably always had enough castings out in their yard for making a run of lathes or other machine tools. I think if you were to call Monarch and ask them to make you a new 10EE toolroom lathe (assuming you had six figures in your checking balance), Monarch would be using castings poured in the 1970's or early 80's. I know when the "Cole" was heavily damaged in Yemen, her machine shop was damaged as well. The Navy put an order in for a new Lodge & Shipley lathe of the same model/capacity as what had been in her original machine shop. The group (Lucas Precision at the time) which owned L & S made a new L & S lathe for the Navy. My guess is they used castings they'd had sitting in the yard from when L & S was producing lathes on a steady basis.
My own engineering opinion is that nothing equals a casting for a machine tool bed. It is my own sense of the dampening and stability inherent in a casting. I think Brown & Sharpe tried building some precision grinder bases or beds in the 1980's using welded fabrication vs casting. B & S was said to have resorted to a vibration process to "align the molecules" for stress relieving. How many grinders were built this way and how they held accuracy is something I never found out.
I've seen the differences between weldments and iron castings on hydroelectric turbines and generators and on other heavy machinery parts. Each has their place. Weldments just do not have the same properties as castings, no matter what advanced design methods or manufacturing processes are used. Once high tensile strength iron came onto the scene, it changed the whole picture for castings. Castings are used where forgings and weldments were previously used. With improved foundry methods and high tensile strength ductile iron, it is often more cost-effective to go to a casting. One pour and a lot of fitup/welding or machine work is eliminated.
I took H-D's factory tour at their York, PA plant in 1984. Interestingly, I had been at the then Allis Chalmers (now Voith) hydro turbine plant for the week previous. A-C was already using robotic welding and robotic finishing (robotics using ari grinders to finish contoured parts of the hydro turbines). H-D was building the frames using humans to put the pieces in jigs and manually weld the parts together (GMAW or MIG). In 2006, I picked up my certification as an American Wleding Society Certified Welding Inspector. One part of the exam was the practical. It required candidates to examine a number of specimen welds and identify the process used, position the weld was run, and direction of travel, aside from identiftying "indications" (porosity types, undercut, underfill, crater, etc). I had always made a game of inspecting any welds I saw, whether on a structure I was walking thru (like an airline terminal or building), or on anything else I came accross. When I got my 2005 HD lowrider, I could see by the welds on the various parts that it was welded automatically. HD came a long ways in 20 years. But, at least on the air cooled engines, H-D was still using cast cylinders with iron sleeves, looking like a sand casting.
Each process has its place and each manufacturing process and material have specific engineering properties. Manual machine tools will likely never be a high production item like Harley Davidson motorcycles (which, as motorcycles go, are are probably have fewer built each year than most other mfrs). The manual lathes left in production around the world seem to have clung to the use of castings, and I think it is a combination of the properties of castings and manufacturing costs. I am an oldtimer in my tastes in machine tools as well as powerplant equipment, locomotives and motorcycles (Airhead BMW's for 40 years, the HD was an unexpected newcomer) and so much else. My opinion is based on over 40 years of experience, but I will also say I am biased and often base my opinion on my own gut or sense of things as much as "running the numbers". One thing I do NOT base my engineering on is cost. I am fortunate to have my niche where I can live by "Heavy is better" and reliability and longevity usually come ahead of first cost. If a powerplant turbine is down on an unscheduled or forced outage, the penalty charges and costs of replacement power rack up so fast that anything, no matter how much more the first cost is, will be thrown at the problem. As long as the bean counters are held at bay (bean counter = some corporate types with MBA degrees who look at financials first and the main objectives second), this kind of engineering flies. Once something blows apart or s--ts the bed during a time of peak power demand, that is when the bean counters get blamed and told that if they listened to the engineers up front, we'd be making money hand over fist instead of paying through the nose and then some. Of course, the bean counters get their bonusses, and the engineers at the plant get penalized for the lost generation.
With machine tools, I think the design philosophy in the USA was to build it to last, and build it rugged. The Chinese seem to have taken the basic designs and leaned them out, doing it on a "thumbnail". The claim is computer modelling allows for lighter "tuned" castings or somesuch. This is probably the bean counters looking to sell machine tools for the highest possible profit. The reason the older classic US Machine tools were so good and lasted so long was due to the fact that no one skimped on the castings. IMO, it is that simple. Rugged castings= rigidity, stability, and good vibration dampening.