Water is heavy. 5000 feet of it is a whole lot of pressure that such a tunnel would have to withstand.
The "simplest solution" is often one that is completely impractical given the facts.
There are few strong currents in the Gulf - it is a source for the Gulf Stream, but internal currents are fairly weak compared with those dealt with regularly in the oceans.
I don't think the solution is practical in this case for various other reasons, but the fact that offshore platforms exist incites me to think that long, underwater pipes can in fact resist ocean currents. They will also be naturally buoyant so long as there's enough air pockets in the material (or just floaters as the other person posted).
In other words, even if it's filled with water at 5000 feet of equal pressure, isn't there still a much greater pressure exerted on the pipe's walls at 5000 feet than there is at 100 feet?
Or am I misunderstanding something fundamental about water pressure?
According to this ( http://www.calctool.org/CALC/other/games/depth_press ) the pressure would be around 15 MPa (this isn't hard to calculate by hand either). Consider that the mud/oil pushes 15 MPa, and water the same: that gives 30 MPa of purely compressive stress. Concrete could take this much ( http://www.cement.org/basics/concretebasics_faqs.asp ), and steel even more easily ( http://en.wikipedia.org/wiki/Tensile_strength -- if I'm not mistaken, steel behaves roughly the same in compression and in tension).
For every complex problem there is an answer that is clear, simple, and wrong.
-- H.L. Mencken
Building things in water is many many times more complicated than the same things not in water. Which is why deep sea drilling is so controversial to begin with.
Flash forward a few crazy ideas (namely tossing a bunch of trash into the BOP to try and clog it) and they went back to a similar method called the LMRP. This time they sawed off the riser from the BOP so that the opening was flush and designed a containment cap that fit better around it than before (less of a chance for water to mix in and make the hydrate crystals) and developed a system that continually pumps hot sea water and methanol down the pipe to also help tap down on the hydrates. This method is actually gathering oil, but since they sawed off the riser a lot more oil started leaking so the net result is even more oil is getting into the ocean. BP hasn't "sealed" the LMRP for fear that it will blow out the BOP, which is pretty much what the Mother Jones piece is about.
tl;dr they have and it's not that easy.
It seems the article was too long for many people's patience/understanding threshold.