The thought process behind the approach was that bitshares as a collection of technologies and businesses may take a long time to fully design and implement and there also wasn't a good test client with rich data about how the system was being used. By releasing a popular, modular piece of software, I3 could grow a user base and start collecting developers to build DACs on top of a common piece of software I3 was responsible for maintaining.
Marketplaces like cointagion (http://www.cointagion.com/
) could be developed easily and even ported into ads, both centralized and decentralized app stores could be developed and thanks to the modular approach, one could also integrate other clients like counterparty and ripple into Keyhotee as an app. New features could be pushed quickly (say on a weekly basis) and bounties/hackathons could constantly add new apps thus growing the I3 userbase.
Bitshares would naturally be the ultimate beneficiary of the strategy as it would be the core application of keyhotee and thus receive the bulk of user attention. The userbase also would give I3 shares tangible value via the notion of customer acquisition cost allowing for investment from traditional VCs.
For better or worse, this strategy was not pursued and it seems Keyhotee as become a minor product in the portfolio.