Your strategy is mostly fine. If a domain pays for its upkeep when parked, it's worth keeping indefinitely.
Coolness factor is also good - but that's subjective and often it's a guessing game. Keyword domains are almost always better, but then again pure keyword dotcoms for reg fee - they are very, very rare if they even exist anymore.
Development is good. However - for development, I always, always go for the pure keyword domains. They are so much easier to get ranked.
I have had some reasonably competitive keywords appear on the first SERP within months after setting up a minisite + some basic SEO, and they have started making significant PPC income pretty quickly. Not huge income, but enough to make we want to keep them unless there's a really good offer.
here's the thing: They have all been pure keyword domains.
Dotcom is king, but I've found that pure keyword names in non-com tlds are often better than a non-premium dotcom.
Example:
http://distancelearningmba.org. I regged it 1 year ago, developed a minisite about 6 months ago at a very low cost, now it's on page 1 for "distance learning mba". Being on page 1 in Google for this business is already enough to make it a $x,xxx domain.
That's the power of pure keyword domains.
My first year of domaining I regged about 600 domains, out of which I sold about 10 (most for low $xxx) and then renewed less than 100...you live and you learn.
Well, you got me going
As for EstiBot, it's often in the right ballpark but of course not always - be critical. I use it myself as that first checkpoint when researching a domain, but take the dollar valuation with a grain of salt, and do a lot of further research.
