A good sign that you have a problem with a project is when the engineers throw everything they have at it. Look at the options for re-inventing 520 in Seattle and you see bigger roadways, new ramps, new draw bridges and overpasses, tunnels, lids, HOV lanes, man-made parks, transit hubs, rail stations, bigger viaducts, massive interchanges, taller and wider structures. It's as if the whole transportation tool-kit is being thrown at the problem. And please, let's be clear, nothing currently proposed for 520 is a simple "replacement." The new bridge, in whatever form, is slated to be bigger and an occasion to deal with other issues, like adding highway lanes or "fixing" bottlenecks at places like Montlake.
This is one of the reasons that "preferred options" and "consensus choices" and "compromises" are often suspect: in trying to please everyone, they please no one, and often do not solve any of the real problems. They too often create a duck-billed-platypus of public projects. Much political weight is given in Seattle to simple agreement, as if miraculously coming to terms on anything ought to be rewarded, but we have a long history of agreeing to do the wrong things, and paying for it later. We let I-5 slice the city in half and have spent multi-millions since trying to repair the damage. Fortunately, we also have a long history of second-guessing, activism, and stopping the "inevitable" in its tracks. The