sorry for the book, but this seems to be turning into a popularity contest rather than an objective look at platforms, I don't claim true objectivity, but here I think is my most rational and objective look at the subject:
vi·a·ble (v-bl) adj.
1. Capable of living, developing, or germinating under favorable conditions.
2. Capable of living outside the uterus. Used of a fetus or newborn.
3. Capable of success or continuing effectiveness; practicable: a viable plan; a viable national economy.
When we talk about a viable platform we are talking about three things I think. Is it developing currently. Can it continue under current conditions indefinitely. And can it fullfill it's purpose capably.
As good as classic was, and I do like it alot, it is seeing some limited current development, it can fullfill its purpose but really only with legacy apps, but it cannot continue indefinitely. We can fix buggy capacitors etc. for only so long...the hardware is dying.
Minimig is a fantasic design that remedies much of the classics drawbacks but isn't really a platform as such, as it is not in any kind of active production, and is not forward looking (I would love it if a company started building minimigs in a600 clone cases)
OS4 is seeing some development in both hardware and software, but the hardware is highly specialized, difficult to find, and very difficult to replace or repair. It is a niche even within Amiga communities, and cannot continue indefinitely due to the limited runs and market penetration of our small community.
The same unfortunately can be said of Morphos, and other custom designs such as Natami. Too expensive, too rare etc.
UAE is a great bit for software that has kept many of us using amigaos and software, but is just that, software, it isn't really a platform unless you consider some livecd type implementations, but noone is really doing that much.
Which leads us to Aros. Many of you know I favor Aros and have said so numerous times, and with reason I think. Aros is under current, significant, development. The software is developed by the Aros team, and literally billions of free (to us anyway) development dollars are pouring into hardware development by companies around the world. Don't like x86 and its memory addressing limitations, bang, 386 etc, fixing all those problems, dont like the limited registers etc. of pentiums, bang, here is 64 bit cpus with more registers. Don't like the limited backwards compatibility of '040 and '060 68k's...well sorry here is a patch. Don't like the expense, and feature shifts of PPC... sorry that sucks.
Aros also doesnt depend of the fickelness of AI, or strength or the Aros team. It has lasted for years with no license, through all the lawsuits and stupidity, and is open source so even if all the current developers were to be abducted by aliens to work on their os, development would continue. Not just on the core os but on the apps that are under development as well.
And its viability as a hardware platform is assured as it is not dependant on small expensive manufacturing runs. We can all try Aros since we all have or can very easily get x86 hardware. If you havent tried Aros, go to the Aros website are try a livecd.
We all have our favorites, we all have ideals, but when you get down to real longterm practicality Aros really is the best, most viable, option. I don't thing that Aros will take over the world, but there are reasons that Linux hasn't. It is too complex, the distros are too different, the community is too splintered. There is no reason Aros couldn't be what Linux tried to be, an alternative desktop os, that is simple, fast, capable, and free.
And that is the most viable amiga option for me Aros (with Uae)