Users who use platform X and browser Y where your app doesn't work will simply switch to another app which does the same thing that was written by someone who _does_ care about all browsers.
Those users won't complain to you, they'll just switch.
If the effort to provide compatibility and an additional set of browser work-arounds exceeds the benefit given the insignificant share of traffic these other browsers provide then it isn't really a loss, is it?
> Those users won't complain to you, they'll just switch.
In my particular case, I only make things that don't exist already i.e. I'm not making yet another twitter-posting app or social bookmarking app. So I highly doubt my target users will switch simply because my app doesn't work nicely on a browser I don't even know about.
Also the solution to this would be to show a small notification if my app is being run in an untested environment and let the user complain with a 1-click form if they wish. I think most people who have a bad experience will let me know that is the case.
That bit of snark ignores the reality that a developer has resource constraints, and they constantly have to decide how best to deploy their time, money and energy.
Failing to address some platforms might well be a big win. Claiming it's a definite loss is absurd.
The original article's point was about progressive enhancement. If you're starting a project that requires, say canvas, it is your loss. If you start with tables and images, the site will work on a lot of browsers.
Those users won't complain to you, they'll just switch.
Your loss.