Most folks already know this, but problems aren't created equal. Even problems that are--all other things being equal--equal really aren't. At least not in terms of getting themselves fixed. In a perfect world, we would just do the right thing out of hand. In a merely improved world, we would soberly weigh the cost-to-benefit ratio with a full accounting of human factors. But in this particular world it so often seems to boil down to who loses the most face by admitting the problem vs. who loses the most face when the problem becomes public knowledge.
In such cases, the surest way to fix a problem that you don't have the tools to fix is to make it the problem of someone who both does have those tools and who also has some skin in the proverbial game. (I can gratefully say that I've never had to work in a situation that's forced me to break something more thoroughly in order to guarantee that it's fixed.)
Finding ways to delegate problems up the food chain rather than downward is, I would argue, among the top ten problem-solving skills you'll ever learn. It's also a good cure for the folks who like to create cozy little sinecures by making themselves "indispensable" (and, inevitably, bottlenecks).
At work, three of the lynch-pin people are out for the whole week, which leaves our alpha-geek plus your faithful blogger to figure out what these folks "just know." And I will have to be out of the office most of Friday as well. I'm looking forward to seeing what procedural changes and mandatory cross-pollination/documentation will come of this.