I'm not saying Unicomp doesn't make a good keyboard. It is a good keyboard in its own right for the money, but there are definite differences that distinguish it from Model M's manufactured under IBM's name that disqualify it from being a replica, in my opinion.
As has been pointed out before, cost cutting with the Model M started back in the 80s. You can tell the difference between a Model M made in the mid 80s and one made in the late 80s/early 90s by weight, force required to actuate, and some of the internal details.
I own a Model M made in 1987 and one made in 1993. Both were manufactured by IBM. I'd bet money you couldn't tell them apart without looking at the label.
You can definitely tell the difference between an early 80s M and a mid-to-late 90s 42H1292 in terms of weight, internals and case details (such as where the cable exits the keyboard and the absence of a speaker grille), actuation force, printing (there were only monochrome legends on the later ones).
Those are design differences, in my opinion. I mean, Unicomp keyboards weigh less and don't exhibit the metallic ping that IBM Model M's do. However, I can't see that being an unintentional decision. In fact, some people don't appreciate the twangy metallic ping, nor want a heavy keyboard. But who would intentionally decide to design a keyboard with key cap flashing, case creaking, cosmetic imperfections on the surface of the plastic, extrusions in the plastic, or case flex? And what consumer would want that? I mean, I can deal with certain design choices, and I can deal with a minor defect. But what we're talking about are major qualitative issues that, collectively, give the impression of a shoddy imitation, in my opinion, but that's me.
Some will say that a Unicomp is an inferior product to the old school Model Ms, some will say that they are more than happy with them and they are just as good as the old IBMs in their opinion. Who is right? As far as I am concerned, it's a matter of taste.
By definition, inferior means "a step down". Thus, unless one simply ignores all of the qualitative issues that I cited, I can't see how anyone could possibly consider a Unicomp anything but inferior to its predecessors.
Some people are perfectly content with their Unicomp, defects and all, and I say more power to them. If you're okay with it, then so be it. I'm not knocking anyone for liking them. I mean, they're not bad keyboards for what they are. Personally, they strike me as imitations or clones more than IBM Model M equivalents, but again, that's just my opinion.
At any rate, I think whether something is a clone or not is dependent on the relationship between that design, and the design of its alleged 'parent', and also legal issues (which are moot in this case because Unicomp legally own the designs and tooling).
To me, it's not that complicated. This isn't a court of law. Words have meaning, and as long as they're used in accordance with the dictionary definition, it's a simple affair. Does their usage conform to the dictionary definition?
I also think that the issue of whether something is a clone or not is largely independent of quality.
By definition, a clone is an imitation, and a replica is a facsimile. Are you really suggesting that the difference between a crappy, hand-drawn imitation of the Mona Lisa and a bona fide replica has nothing to do with quality? Really? So, a version of the Mona Lisa scribbled out by a 5 year old wielding crayons wouldn't be considered an imitation(ie. clone)?
With respect to B) this has also always been an issue, as it will be with anything that is mass produced over a long period of time. One example of this is keycap printing, even back to the Model F days, you can see some keyboards with poor or badly aligned printing. One of the guys at Unicomp mentioned that every once in a while, the machinery used in the dye sub process would need to be retuned, and that some keyboards would go out with fuzzy or unaligned legends before they'd fix the process up (this dated back to the IBM days) Again, I don't see how or why failure to retool your manufacturing machinery suddenly turns your products from the real deal into a clone.
If we're talking about a single qualitative error that's handled from time to time, well, that's one thing. But that's not what we're talking about. We're talking about multiple major qualitative issues that are consistent and ongoing.
If I go out and buy a Filco tomorrow, it is not a 'replica' of the Filco that someone bought last week, it's just another Filco.
If you purchased the same model, assuming there were no manufacturing errors, it could most certainly be considered a replica. What do you think a replica is? It's a duplicate, facsimile, or reproduction.
Given that the Customizer/Classic 101 is built on the same machinery as Lexmark built 42H1292-type Model Ms for IBM by former Lexmark employees according to IBM/Lexmark designs and specs, a Unicomp is just another Model M.
I defer to my McDonalds Big Mac analogy:
Let's say a company, we'll call them Uniburger, bought the rights and equipment to make Big Macs from McDonalds(because McDonalds planned to stop selling them). And let's say Uniburger made a decent burger, and even cooked their burgers on the same equipment that the original Big Macs were cooked. However, let's also say the quality of Uniburger's version of the Big Mac was clearly inferior to the original. Do you think most consumers would consider Uniburger's version to be an original Big Mac or an imitation?
Pretty much all the differences are in line with something that has been built for nearly 30 years
The quality control issues are not at all consistent with what was produced by IBM or Lexmark.
