Just be careful what generalizations you draw. I've been in enough proposal development meetings to know that the estimates that make it to the client/government entity aren't always the actual estimate. In many cases, one or more persons provided a pretty good estimate and it got shot down immediately because if the proposal included it, it would be rejected. Basically, if there's a good chance the client will eat the cost overruns (as governments usually do for military/space/infrastructure projects), there's a big incentive to deliberately underestimate even when you know the actual number will be significantly higher.
I was about to say, estimates better be aligned to the sales team expectations or they'll be changed so that they do align. It's only after the project is sold do you get a chance to look at the reality of the situation.
