What have I worked on since: ACDZip, ACDSee, Tracktion, EAW Resolution, SPIN Review. ACDSee was a brief success, but the rest - meh.
This magic-bullet interview question had me depressed and made me feel like a failure. I too have nearly 15 years of professional experience. My last projects that I can show off were 4 and 8 years ago.
I don't feel like a failure anymore. I got a reminder that success depends on blind luck a lot and is rather overrated. Hard work, improvement and patience are still essential.
P.S. I have used both Miranda and ACDSee in the past. They reached my highest criteria for software, that few applications reach – were small, fast, reliable and very useful. I wish FigBug best of luck.
Any given project will be calibrated somewhere between the extremes of "just ship it" and "take 5 years to make a perfect 1.0". A developer who, based on experience, is flexible enough to calibrate themselves to match the project -- who can adapt to reasonable balances of "do it faster" vs. "do it better" -- is valuable.