That's actually exactly what they're doing, but slightly different approaches.

Patent:
Uses adaptive impedance matching to allow signal propagation through the plasma.

This project:
Uses adaptive frequency matching to allow signal re-transmission through the plasma.

Same net effect, exploiting the same properties (The patent changes the impedance of the transmission circuit to match the plasma, while TFA describes varying the frequency until we hit an impedance match with the plasma (changing the frequency changes the effective impedance of the plasma).

Increased memory usage over time is often not a memory leak, and not always a bug.

If the memory usage increases without bound and begins to conflict with other programs that are attempting to use memory, then it's a bug. Anything else may just be good caching behavior (If you can cache something to improve performance, that's better than leaving memory underutilized).

Also, some sites will load slowly because of google-analytics. Are they going to be penalized by Google, or will they be considered "faster" than other sites that aren't "enhanced" by Google services? So, a site that takes 6 seconds to load, with 4 seconds waiting on google-analytics is slower to a user who's not blocking google-analytics than a site that takes 3 seconds to load. Google will probably compare the 2 second time to the 3 seconds, and rank the Google customer site higher than the faster third party.

MUMPS: w 42
If you want to clear the screen and add a new line: w #,42,!

The teflon covers the semiconductor - idea is that if the metal has an irregular surface (which it does), the semicon presents a regular surface to the conductor and insulator so there are no E-field 'hotspots' that can break down insulators very quickly.

Almost -- has to do with safely dissipating the electric field. In buried transmission cables, losses are minimal (they are coaxial, so there's very little radiation of any kind), but they use an engineered layer of semiconductor between the conductor & the neutral sheath to create an E-field gradient. On a very high voltage line (say, 169kV & higher) this semicon becomes impractically thick (not to mention extremely susceptible to failure). When you can achieve the same thing with 20 ft of air, which is almost free (you have to buy land), and maintenance costs drop significantly (for not having to dig), it just makes sense to do overhead lines.

Haven't you ever heard of "good enough for government work"?
It *was* good enough, but... um... well, that guy over there who I've never seen before mismanaged the project, and that's why it failed!

From TFA:

School officials have admitted that thousands of students, faculty and employees could have accessed the same file for up to two weeks

So, thousands of people have had access to this file, and the one person who tried to report it (and was tracked down) is being charged with felony counts of computer access and identity theft? And they're not checking to see if anybody else has tried to access this file, to indict them, as well? Definitely seems like a case of shoot the messenger. According to a state trooper interviewed in TFA,

He deceitfully used someone else's name and password so he would not get caught and was looking to profit from his criminal act.

I didn't see anything about him trying to profit, though... He sent an email to the principal (contents unknown), from an anonymous email address, signed 'A Student'. Without more info, I'm inclined to speculate that he didn't really appear to be attempting to profit. (Wouldn't it be better to keep this a secret and profit from the information, if that was really his intent?)