The app appears non-trivial to the user. There is a QR code reader that takes snapshots of industrial equipment, saves the URLs to a local database, long-polls the URLs to continually display fancy gauges that mirror actual hardware output in real time, and pops up service information when equipment stops running.
Performance-wise, the app is laggy but the lagginess is perceptually the same as similar well-established native apps. All QR/barcode scanners are slow and all data access over wireless is slow so what if Titanium tacks on an extra second or two. (I don't have any benchmarks either way, it could be the same for all I know).
To explain, I believe it's an established fact that to the end user, anything under 2 seconds is "fast" and anything over is "slow". There is no differentiation.
Productivity-wise, the app took 2 full days and several evenings to develop. For reference, my programming background is: web development - advanced, javascript - fair but hardly advanced, mobile - zero, stateful/compiled/memory managed applications - zero. As you can imagine, Titanium takes all the credit.
My overall impression is that Titanium is the right tool for the right job (horrible documentation notwithstanding). The lack of "awesome" apps is probably because it's not right for games and hardware-centered apps. To put things in perspective, I am a medium-heavy app consumer and for me personally, games and hardware-centered apps comprise maybe 25% of what's in my iPhone and iPad. That leaves a whole lot of applications that could easily be done with Titanium.
Go to the Apple's marketplace right now and filter out all the games. Then filter out all the apps where the app is just a thin frame for the content (I say that because I am pretty sure you wouldn't notice if Titanium was used in place of native, because of the 1-to-1 API to native UI). I bet you are going to feel that there are very few awesome showcase apps left there too. That's just the nature of the beast.
Yes, limitations in our abstraction layer is the biggest engineering challenge we face. That's true of any framework that creates a higher level API over the top of an existing API (Rails, Struts, Flex, etc). As the abstraction layer hardens, hardware gets faster, and platforms improve (Android is, IMHO, an inferior platform in terms of UI performance generally, but is getting better)... our performance across the board will improve. It's very usable and widely deployed now, but we know it still needs to get much better.
main site: http://www.6wunderkinder.com/wunderlist/
One caveat, I'm not sure how many of the current apps (there are 5 versions with an addtional blackberry app in private beta.) still uses Titanium, and Wunderlist has been ported to platforms that are not supported by Titanium. However I know that the first few releases did use Titanium.
I think the IOS version is still Titanium?
A high performing Titanium example on android would interest me also.
ZipCars pretty notable...
The main concern I have about either of these is that there are no awesome ahowcase apps out there that prove they work well. There are quite a few awful ones. I am attracted to titanium's promise of native look and feel, but the author admits that there are unresolved performance issues and other drawbacks. Phonegap seems even worse when compared to native UI.
What is the best nontrivial, well-performing titianium or phonegap app you know of?