Well, according to http://www.paulgraham.com/startupideas.html you can literally go and find good ideas right now and every day thereafter.
It is also the basis of the Lean Startup Machine weekend (phenomenally popular).
You can train yourself to have ideas. Good is just an evaluation system.
People may indeed be incorporating a new product into their lives and the Founder will need to manage customer expectations and fallout. A high likelihood of discontinuance does not disqualify his attempts from being a startup.
The truth is, from a validation perspective the OP may have stumbled onto a supremely successful model of idea validation. Build it, tweak it, pump it for 11 months. The next project has 10 months. The next 9. Etc. Compare the metrics after one year and choose the most successful.
However, the statement that good ideas do not come up routinely and is false. They may just not be routine for you