http://www.baseballprospectus.com/article.php?articleid=2844
http://www.baseballprospectus.com/article.php?articleid=2851
http://www.baseballprospectus.com/article.php?articleid=2869

Click does a very thorough job of assessing when bunting improves your chances of getting a run, taking into account the skill level of the hitters who follow the bunter.  I recommend the series to you.  If you are not a BP subscriber (and you should be), then I will quote from the conclusions:

"When run maximization is paramount (early in the game, in high run-scoring environments, etc):

• Only pitchers should sacrifice a man from first to second in any circumstances. Even then, certain pitchers who are decent hitters should swing away.
• With a runner on second and no one out, sacrificing makes sense when some of the league's worst hitters are due up, with a hitter with a high propensity for singles and doubles following. The most likely instance of this is as a lineup in the AL turns over from the ninth spot to the first spot. Even then, instances where sacrificing increases run expectation are rare.
• Sacrificing with men on first and second is only a good idea when pitchers are due up. While the thresholds here are higher than in Situations 1 and 2, they still remain far too low for even the worst regular position players.

• Similar to the run maximization situation, only pitchers should sacrifice a man from first. Given that a pitcher would likely rarely be batting in this situation where runs are at a premium, this situation is likely to never occur.
• Most of the league should sacrifice a man from second with no one out. While a line of .277/.350/.451 is slightly above average, recall that the skill set of the second batter due up should also be taken into account. On the whole, this finding is in the greatest agreement with conventional strategy.
• When runners are on first and second, sacrificing is, again, not a good idea, a finding that is due almost entirely to the opposing manager's propensity to intentionally walk the next batter to keep the double play in order. This 10% decrease (approximately) in the scoring probability of the situation is enough to reduce the threshold across a great deal of current hitters.

Therefore, in the broadest conclusion possible, we can say that sacrificing is a good idea when pitchers are batting and, for most of the hitters in the league, when there is a man on second, no one out, and a single run is the goal. Even then, there is a set of the league's best hitters who should never lay down a bunt; which is too bad, because it would be fun to see Bonds square around, just once."

As to your main question about the disruptive effect of bunts, I'm not aware of any definitive answer.  I seem to recall that someone did some work that concluded that if you include the chance of bunt base hits and errors into the calculus, that did make sacrifice bunting somewhat more palatable.  However, if you are going to make that argument, you must also take into account all the times that a sacrifice bunt fails, because the lead runner gets thrown out, or the bunt is popped up, or (horror of horrors) there is a double play, or even when the hitter puts himself in a two-strike hole failing to get the bunt down and thus lowers his chance of success when he finally swings away.  My entirely unfounded guess is that the bad unintended consequences of sacrifice bunting are about equal to the good ones but I'll root around a little and see if I can find any real evidence.

I also think optimistically that there are many examples of how objective, quant-based analysis has improved the game.  Pitch counts for example.  When I was a kid, no one paid the slightest attention to pitch counts.  Now, you may think we've gone too far in babying pitchers (and you might well be right), but the fact is that countless great arms have been lost to thoughtless handling, Sandy Koufax for one.  Wouldn't it have been great if Koufax had been handled gently and pitched for another five years?  Conversely, wouldn't it have been rotten if Pedro Martinez had been ridden like Koufax and flamed out at 30?

While I agree that evidence can be unmeasurable (and I'm not sure what you mean by "transient" in this context), the problem with non-quantifiable evidence is that it so easily turns into self-selected, self-justifying anecdote which is worse than useless for making a decision.  For example, when Hawk goes on about how great it is for Iguchi to give himself up, he only mentions the times it leads to a run.  He doesn't mention all the times if fails or the (probably unknowable) loss of what Iguchi would do if allowed to swing away.  When DJ prattles on about how "aggressive" base-running elevates the team, he never aknowledges the times aggression takes us out of innings.  

So what I'm trying to say, however haltingly, is that there is clearly evidentiary value in the non-quantifiable observations of smart, experienced people (and probably other sources as well if I think about it) but that kind of evidence needs to be handled with extreme care because so often it is nothing more than anecdote that has been filtered through the assumptions of the observer.  Ideally, I'd like to be able to test that kind of evidence in a measurable way.  For example, I'm quite willing to stipulate that Leo Mazzone's skills and experience help pitchers succeed.  I don't know exactly how and I certainly couldn't quantify his methods but I can look at the results and be confident of the conclusion.  Conversely, I have no reason to believe that Pods' aggressiveness makes his team-mates better, even though experienced baseball men keep telling me it is so.