"No options for lower Res screen, so this app isn't an option for me."
"Has potential, but doesn't resize for smaller screens, so can't see half of what's going on. Interface is snappy though, which is good."
"screen resolution issues: Doesn't display properly on my Samsung Galaxy Ace. Hope you can fix this. Otherwise, quite an interesting app."
"My phone screen is too small for this app"
Along with a slew of other problems with bugginess, lack of feature parity with iOS, etc. Telling is the variance of its score (http://cl.ly/0m3b2a2v3s1Y0x071Q0e) versus iOS (http://cl.ly/030I0M3z1I181g3T2p2T), suggesting that the Path developers have this product working well on some phones, but terribly on others. That's one of the clearest examples of fragmentation that I can find.
If I buy an Android Device (tm) and get something that has a (both in terms of visuals and functionality) botched and outdated version of the Android project, I care.
If that software will be forever behind the platform and _not_ as you state it 'work perfectly fine', I care again.
For me this was a very real scenario with the HTC Hero (no idea what the name outside of the EU was). I preordered the phone, got a brick with an old version of Android with a non-standard ui and while the world moved on I saw more and more applications on the market that didn't like to support my 1.5/1.6 device at that time.
I claim it's better now, but mostly because Google took a lot of time between the last handset release and now ICS. Phones could catch up to 2.x, crawling sloooowly from update to update, while Google didn't release 3.0 and worked on ICS for a while. I'm pretty sure we'll see a great gap in adoption again very soon: The manufacturer are going to drop support for ~most~ handsets or take month to create an update for existing devices.
The main problems I see with this is (apart from people like me that feel left behind/screwed if you sell them a device that is more or less DOA) that the "brand" loses a lot of value. If you compare a shiny "Android phone" with another "Android phone" and one looks gorgeous and runs fast while the other looks wildly different and is slow as molasses you're alienating the customer. You force your customers to do a lot of "search the reviews" dances before buying a mobile or they end up with a POS that is never going to do what you see on "other Android phones" around you.
With that in mind, the final product will get a lower quality if the developer can't afford the costs of optimizing for every device. Or your device incapable of running the app at all.
> Or your device incapable of running the app at all.
If your device isn't capable of running the app, it won't install (or even appear in the Market for that matter!) assuming the developers aren't entirely incompetent.
Credentials: I'm an Android developer.
It was JavaME's greatest nightmare (one could never be sure an app would work at all on different devices). Luckily, since there's no JCP and vendor-specific packages, at least that part of the fragmentation was solved...