I initially upgraded to testing in order to gain hardware support for my touchscreen, wifi, bluetooth and graphics. However testing does come with problems, and many new users don't quite realize what those problems mean or assume that won't happen to them. Quite often new users either run testing, or even worse, a mix of debian testing and stable, break something, and come here asking for help when the best advice is to stay with stable and use backports if they really need something more up to date.
Today for me is a good example of why, even though I'm going to continue using stretch for the time being, I plan on following it into stable and staying there once it's released, and next time I have hardware problems I'm going to spend some more time trying to make stable with backports work, even though I wouldn't call myself a new user.
I'm not quite sure what update did it, but sometime in the last few days both google chrome (for netflix) and my main browser, chromium, stopped working, passing segfaults. gdb told me that '/usr/bin/chromium' is not an executable file. When I went to reinstall, apt-listbugs tell me that a major bug as been filed against chromium that the current version crashes immediately (so it's not just me! yay!). This is a problem that would not be encountered in stable. There is apparently a fix in sid, but given how testing works it'll be at least 10 days before I see that fix, so it's firefox for me until then. Just thought I'd share in case anyone finds it an decides to stick with stable.
Testing is not stable, even if it works fine the major of the time, because s**t happens and it happens on testing far more often than stable. It's not even as stable as a *buntu, except during freeze, because updates arrive just a few weeks after they are released (example: I am using Gnome 3.20, Ubuntu Gnome is *mostly* using 3.18) and clearly some bug in chromium (or in another package that effects chromium) made it through 10 days in sid and landed in testing.