This is the thesis of my argument. The governance is too inert, closed, and rigid, and increasingly, due to the game structure in the witness voting, it becomes even more inert, because of people getting fat on its corruption.
Yes, absolutely, more important than anything else for a platform like Steem, is effective and open governance, that is widely participated in. The communication between users and developers is such a tenuously narrow channel that they are really missing everything that is not the 100% flavour of the month topic of conversation. Every smaller issue gets drowned out.
Probably, therefore, number 1 priority of the DIV project is to ensure that in parallel to the general forum, there is a monetary incentive based code governance system that factors in reputation scores within its boundaries, to ensure that bug fixing, feature requests and implementations, and hard forks, are made simpler, more tightly scoped, and less political, and more utilitarian.
There also needs to be a 'sandbox' where there is no way to exchange tokens, because it is frequently reset to zero, where significant new changes are tested upon parallel contemporary content, with a different set of votes applied, or some similar sandbox. This way at least some of the surprises we have had to deal with don't directly impact our bank balance.