I have reported about 8 or so in the last year. A few have been fixed (usually as a result of several other people reporting they were having the same issue). I had to fix a couple of bugs myself. The other half were never fixed. One I reported in a Bugzilla type database and within 24 hours it was marked as "Closed" without any comment from a developer. The next release of the software had the same bug. One I reported on a forum for a closed-source application and immediate a dozen or so other users agreed with me (it was a memory leak, causing a background daemon to consume 10-15MB of additional RAM each day it continued to run). The company representative said that they could not reproduce the problem.
Most recently I reported a bug in an exercise tracking piece of shareware. The software imports data from Garmin's software and is able to do a lot more with it. Due to what the shareware developer sees as a bug in the Garmin software, distance for a given activity might change a bit on import. This is fine, and the developer goes to great lengths to explain the discrepancy and why he believes his calculations are more precise. I agree with him and continue to use the software. I eventually realize, however, that while the distance changes and the activity duration (time) stays constant on import, the pace for the activity does not agree with the distance/time. I report this and the developer responds that his software trusts and uses the pace value passed on input, and says that users would bother him to see why the pace value does not match what is in the Garmin software. He explains a way to change the value manually and marks the problem "Resolved".
Note that one of my activities was off by over 40 seconds per mile. What should have been a 6:57 min/mile pace was marked as a 7:49 min/mile. This is a very large discrepancy.
Not very reassuring.