I would just point out that a column of Earth's atmosphere approximately 100 km high exerts a pressure of 14.7 pounds per square inch at sea level. Using gas pressure to move things up and down in a tube elevator entails pressures much higher than this because there is a load at top, and the column is much higher. Also, from an engineering point of view a tube is a structurally efficient structure as long as the curvature of the wall is high--i.e., the tube has a small diameter. Granted the aspect ratio is very small (tube diameter to length) however, the structure's ability to resist pressure is a function of the tensile strength of the construction material, and the amount of tensile load (or strain) placed on that structure, and is practically independent of the length. Tensile load is a function of the internal pressure (gas pressure) being resisted by the structure combined with the structure's ability to resist that pressure (which is a function of structure's curvature) and the thickness of the wall material. The short answer is, when structures become large, then their curvature becomes low, and wall thikness must increase to resist the same forces: the structure becomes less efficient. The increased mass complicates the support structures, so the structural dimensions are complicated again because now the length of the structure becomes important.

This is why I think the proposed ribbon cables are actually a good idea. They minimize the load into a purely tensile load which is much easier to assertain and engineer for. Also, complications will arise if graphite fiber or 'buckytube fibers' are used, because as the structure climbs out of the Earth's atmosphere then atomic oxygen which is far more prevalent at 100 km-200 km, will bombard and oxidize the carbon structure. Erosion by atomic oxygen is a real concern, and coatings of silicon-carbide will probably have to be applied almost continuosly (an interesting parallel to the perpetual painting efforts of the San Francisco Golden Gate Bridge!)

I have often thought that a space elevator concept will only be practical if they were deployed from a synchronous ring structure completely girdling the Earth at the equator. A ring structure could be built as a network of orbital stations building outwards, just as some super long span bridges are built today. When the structures meet and all orbital stations are structurally connected, then the whole ring station could be despun to synchronize with Earth's rotation. The whole structure becomes a self supporting arch that spans the sky!

Space elevators could then be easily deployed wherever one chose, almost trivially by dropping a line down. A trip to deep space would be a matter of catching an elevator up to a Ringway station, and catching a ride on a Circumferencial accelerator--a 'linear' accelerator that accelerates a payload back up to orbital or even escape velocity and flings it away on a tangent. Arrivals could be decelerated similarly, but this could prove to be a bit trickier. It might be easier to capture a returning flight by docking with another structure some thousands of miles farther out that is attached by another cable to the ring station. A docked payload is then shuttled back down the Ringway by winching it downward...something like that.

Honestly though, construction of a Ringstation assumes that nearly all of the problems associated with bulk transportation of raw materials in deep space (asteroid and lunar derived materials) has been completely solved. So the whole purpose of the station kind of becomes a obsolete...

Still, it would be really cool to build one, even if it were millions of times larger than any engineering project ever done by humans.

And it would be something that extra-terrestrials could see with a large array of optical telescopes. Infact, a ring station is a rather bold proclamation of intelligence--it might not be a bad idea for use to optically look for our neighbor's Ring Stations...but that's a whole different story...