Methinks Vancbiker doth protest too much.
The OP proposes to use readily available, relatively inexpensive off the shelf components as the basis for a special purpose, essentially one job, machine. Where the parts come from and what they were originally intended for is of no matter so long as they are up for the job. Obviously a complete Shereline or Taig is completely inappropriate but if the headstock and bed will provide an adequate chassis for the special bits then its stupid to waste time, money and effort re-inventing the wheel.
The issue is whether the parts concerned are up for the job and, if not, what is a similarly economical source for something that is.
Reminds me of the discussions when we were looking into introducing COTS (Commercial Off the Shelf) equipment as an inexpensive replacements for specialist military equipment.
Historically jobs like Automation is interested in were frequently done on capstan and lever collet fitted bench top machines like the Pultra, its bigger brother the Smart & Brown L series or small autos like the ingenious (and arm aching) Britan. Many breeds of such machines were made and the number of parts made on them over the years is well beyond convenient or even inconvenient counting. I'd be looking at such machines for inspiration.
I've long felt that the ability of modern production technology to mass produce machine components of sufficiently decent precision at reasonable prices has made it possible to consider several small, relatively inexpensive, slower running spindles as a viable alternative to a swiss or similar machine running at 96 to the dozen.
For suitably small work of course.
I've amused myself for several years sketching out several variations of what I call my "credit card lathe". Basically cheap enough to go on a credit card, £1,000 - £3,000 maybe, small enough to go in a rack and slow enough not to need exotic tooling. If it lasts a year its more than paid for itself. Want more capacity just buy another machine and stack it on the next shelf up. Bar handling for a vertical rack full is the problematic issue but I don't see anything too hard there. If a weight and chain worked for grandad...
My current favourite concept for baby work is Britan style reworked as a sliding head machine with the rotating tooling carrier fixed and vertical acting cut off slide. Control concept is basically that of a plugboard auto, drive to a stop, rather than sophisticated swiss. Sliding head makes it easy to release the collet by pulling back against a stop to drive the release. Simple bar feed against a stop plate that swings away before the next head stroke.
Around 3 minutes cycle time would get 3 K parts per month. A swiss would die laughing at the idea that such speeds could be acceptable but perfectly practical cheaply, probably 1 minute cycle time is in reach before you have to get serious about machine engineering.
Clive