One major reason it wasn't acceptable to require Javascript many years ago is that many people had it disabled for good reason. Malicious code, popups, and browser crashes were quite common. Furthermore, browser security settings didn't always allow users to enable javascript site-by-site, so enabling it to have a good user experience on your site meant they'd also have a very bad user experience elsewhere on the web.
Most of those issues have gone away. It's still better to have a slightly degraded user experience for your non-javascript users, rather than a complete site failure, but it's not nearly as important as it was in 1997.
Umm... malicious code, popups and browser crashes/attacks are still common, and still rely on Javascript. It has become much less convenient to disable it, but...
Crashes are now mostly the domain of Flash, though. JS hangs in an infinite loop sometimes, but that's not a crash. Actual JS-originated crashes are exceedingly rare in my experience, and are usually the mark of an experimental-build browser.
Popups are largely handled by every major browser having popup blocking that works reliably enough.
Malicious code is what the ad-supported internet runs on. The only thing they can hit is your already-public data - the use of that data is the only thing that can be construed as malicious.
It's still better to have a slightly degraded user experience for your non-JavaScript users, rather than a complete site failure
I would suggest that a company look at their web analytical and make that decision. I know that for most of my clients their trends show less than 1% for JavaScript disabled. This, weighed against maintaining a degrading site programming model, to support 1%, did not make economic sense, given that those dollars could be used for multilingual translation, (a far larger segment).
This is really a question that is best answered by looking at your sites trends. For us it made more economic sense to do a JavaScript detect and if they did not have it enabled pop a page suggesting FF, Safari or Chrome.
I would suggest that a company look at their web analytical and make that decision. I know that for most of my clients their trends show less than 1% for JavaScript disabled.
If you do take this approach, remember to be a good scientist by making sure you respect the control aspect of the scientific method. Specifically, if you already require JavaScript for your site, you should not even venture to extrapolate an answer, even if you think you're going about it in a reasonable way.
I recently had a conversation (although it was less like a conversation and more like a few frustrated, snarky comments from me) with a developer who claimed that 99% of his users browsed his page with at least 1024 pixels. Even ignoring for the moment the issue of screen width not being the same as the size of the browser window on the screen, I asked why he would expect otherwise from users of a web site which required that kind of width before it begins to invoke frustrating things like horizontal scrolling.
Most of those issues have gone away. It's still better to have a slightly degraded user experience for your non-javascript users, rather than a complete site failure, but it's not nearly as important as it was in 1997.