Don't forget, the Osbourne 1 sold many units with a display that only showed 40 columns.
The Osborne showed 52 columns (without the 80 column card), but it was a moving window on a hardware virtual 128x32 screen, so 80 column software didn't require any hacking to work with it. The Osborne also had a large library of well known business software that ran on it. Commodore might have had reasonably good intentions in trying to provide a functional suite of "business" software in ROM, but realistically you're just not going to be able to stuff something as functional as Wordstar, Supercalc, and dBase II into a reasonable amount of ROM memory.
(And let's be clear here, the software built into the Plus/4 was not even *remotely* like Wordstar; the word processor could handle single-spaced documents about a page and a half long...)
The portable office machines all had small screens. Model 100 was 40x8; Epson HX-20 was 20x4. There were potential markets for what Commodore was developing but that was not what the final systems hit.
Any TED portable would have to be a "sewing machine" like the SX-64. (Or be seriously compromised by using one of those tiny "folded" monochrome CRTs... or need an LCD capable of TV resolutions and refresh, which opens up a whole can of worms.) It's not realistic to think of TED *ever* being incorporated into something like a Model 100.
The LCD is a different kettle of fish here; Commodore had apparently invested a fair amount of money in getting a good quality 80x16 LCD working, and if Commodore had managed to get it out the door by early 1985 with a price tag competitive with machines like the Model 100
there may have been a window of possible success, but realistically I think history argues against it. There were a whole grab-bag of similar 80 column portables, like the Epson PX-8 and NEC PC-8401/Starlet, that never sold in anything like the numbers the Model 100 did. In part it was price (these machines were a few hundred bucks more than the Model 100, for roughly similar capability other than the screen size), and in part because with the tech of the time 80 column LCDs were a *lot* harder to read, even if they were "good quality" for the time. And of course the other factor was it was obvious that PC compatible notebooks that weren't going to cost *that* much more were right around the corner. (The NEC Starlet premered in early 1985 for $999; later that year the Toshiba T1100 was out for $1900. Yes, that's almost twice as much, but it had a full desktop-size screen, a floppy drive, and PC compatibility. It weighed twice as much, but if you were carting along a separate storage device for the Starlet instead of relying on its RAMdrive it was about a break-even.)
In short, with the LCD Commodore made the rare wise decision of realizing they'd completely missed their shot and backed out before turning out enough units to dump in Hungary or whatever. TED really should have stopped with the 116; if they really wanted to commit to trying to dent Sinclair maybe they could have chucked out a 64K version of it to compete with the 48K Spectrum, but only if such a machine was in fact materially cheaper to build than the C64. (It was never clear that the Plus/4 was, despite its marginally lower chip count. And it is marginal, because adding the ACIA back in to have a user port pretty much narrows the difference down to the SID and a couple at-this-point very cheap color SRAMs.)
"Couldn't do better" is a huge window though.
Pretty much any 80-column machine (Apple II or an IBM compatible) in the early 1980s is going to cost you over $1000, and add a couple hundred for the monitor. Being able to say "you can bring this thing home for $200 and hook it to your TV" is a compelling case, even if the customer comes back to buy a $300 disc or $75 tape drive next week.
Thing is, though, is by 1982 (maybe '83 at the latest) this fact that a "business computer" required this level of investment was reasonably common knowledge. The "bring it home for $200" game works when you're appealing to parents to percieve your offering as a smarter, more educational alternative to a video game console, but the backlash against this was already building even in that realm. (Everyone knows about the "Video Game crash of 1983", but very much hot on its heels followed a downturn in the home computer market, in large part driven by disillusion with the true costs and limitations of cheap 40 column computers.) A 40 column toy on a TV set *can* do word processing and spreadsheets, but not well. I suppose what's really hilarious about the Plus/4 is that built-in software in ROM, as bad as it is, needs a disk drive to save and load files, so the argument that you could buy a Plus/4 for its $299 price and do *anything* meaningful with it goes right out the window; it's a $600+ machine right out of the box. In 1984 mail-order vendors would sell you an Apple IIe with two disk drives and a mono monitor for $1600; similar money would get you a Kaypro II. Yes, that's "a lot more money", but you're getting *worlds* more capability.
When you itemize what you get with it it's obvious the "toy" $299 computer actually isn't that great of a deal, and people who needed a computer for business were well aware of this by the time the TED machines actually shipped. Commodore should have known this.
They were putting out a huge range of feelers with stuff like CBM-II and TED machines, but I wonder if that was a product of the time. I'm not sure there was a sensible play for a "long-term" home computing platform in 1982 or 1983. If you built an 8-bit machine, it has a bumpy upgrade path (see: C128 or even the Apple IIgs), and a 16-bit machine was too expensive to get into the home mass market, especially when the big takeaway from the success of the VIC 20 was "if you make a cheap machine that's marginally more competent than Sinclair, you'll sell a bazillion of them."
Sure... and I think the question that's reasonable to ask is why in the world did they think building a million different mutually-incompatible machines was ever going to work. I think maybe the best explanation might lie in that quote that
@DrAM19 threw out that talks about Jack Tramiel essentially thinking he was in the calculator business. If you're churning out calculators then, sure, the idea of slapping different function sets and extra buttons on otherwise similar models and saying "this one's a
Business calculator, and this one is a
Scientific calculator..." makes sense; maybe the execs at Commodore just couldn't grasp the idea that computers didn't work that way and trying to stuff them into that model was just creating a huge support and marketing nightmare. Broadly speaking, sure, you can target market segments based on price, but broadly speaking I think it's clear that companies that responded to market segmentation by creating mutually incompatible lines of machines did worse than those that kept to a unified upwardly-compatible model, and Commodore between 1982 and 1985 was the absolute poster child for doing it the wrong way. If they hadn't had that one-hit wonder with the C64 they'd have completely imploded.
When it comes to the CBM-II line in particular I think it's notable how little Commodore learned from the debacle that was the Apple ///, which was very public by the time the CBM-II machines were in development. In some ways the CBM-II line was actually worse, because at least the Apple /// both had backwards compatibility mode with their previous most popular model of machine and offered some significant functionality improvements in its native mode beyond just extra memory. CBM-II machines had pretty lousy compatibility with the PETs they were intended to replace and didn't even compensate for it by having basic improvements like high-resolution graphics built in. It's kind of baffling to even figure out who they thought they were competing with. The TED at least makes sense in *theory* as a cheap-and-cheerful low-end solution for computer literacy and videotext terminals... but not even remotely as a business machine.
