Oh I wasn't even thinking about those things, but that's really cool! I was thinking more about different CSS implementation behavior, though that's less of a mobile-specific issue and more of a "every browser is different" issue.
I don't have a list of differences off the top of my head, but I regularly find big enough differences that I don't rely on "responsive mode" or even (as mentioned) Apple's iOS simulator (because it does not accurately replicate the real on-iPhone browser rendering, which has bitten me before).
Using Polypane doesn't mean you can skip out on testing other rendering engines. That's such an important point that I even mention it on the homepage!
So yeah, you should be testing (mobile) Safari and Firefox too. Chrome on android has some different APIs compared to Chrome on desktop, but rendering is identical.
When developing your site, you can use Polypane Portal[1] to tunnel your local site to real devices while keeping them fully in sync with what you do in Polypane, so you can scroll, interact, inspect and even edit across real devices from inside Polypane, saving you a ton of time.
This reasoning misses a very large portion of Polypane, which is that you use it while developing, not just testing. It's a replacement for your time in Chrome/Firefox/Safari, not your time testing across devices (real or through online tooling like Browserstack).
It does however minimise that device testing time. So it all depends on how valuable that time is for you.
I don't have a list of differences off the top of my head, but I regularly find big enough differences that I don't rely on "responsive mode" or even (as mentioned) Apple's iOS simulator (because it does not accurately replicate the real on-iPhone browser rendering, which has bitten me before).