The Videocore IV on the Raspberry Pi (which totally kicks arse, BTW, it's a beautiful, beautiful processor. Did you know it's dual core?) currently doesn't have an open source compiler that's any good[*] which I'm aware of. I have tried porting gcc, and got a reasonable way into it, but ground to a halt because gcc. I know another guy who's similarly about half-way through an LLVM port. And Volker Barthelmann's excellent vbcc compiler has a VC4 prototype which makes superb code, but that's not open source at all.
Without a compiler, obviously the source isn't much good, although the VC4-specific code is really interesting to look through.
In addition, having done a really brief scan of the docs they've released, this isn't what the article's implying: what we've got here looks like the architecture manual for the QPU and the 3D engine. The QPU is the shader engine. Don't get me wrong, this is awesome, and will allow people to do stuff like compile their own shaders and do an OpenCL port, but I haven't seen any documentation relating to the VideoCore IV processor. The binary blob everyone complains about runs on that.
It does looks like the source dump contains a huge pile of stuff for the VC4, so maybe they're going to release more later. But even incomplete, this is a great step forward, and much kudos to Broadcom for doing this.
[*] I have done a really crap port of the Amsterdam Compiler Kit compiler for the VC4. It generates terrible, terrible code, but I have got stuff running on the Raspberry Pi bare metal. It's all rather ground to a halt because there's still a lot of stuff to figure out in the boot process, but interested parties may wish to visit http://cowlark.com/piface.
...actually, I misread the chart. GEO is 3.8km/s from LEO. Lunar orbit is 2.4km/s if you transfer from GEO, giving a total of 6.2km/s. If you go straight from LEO to lunar orbit, it's only 4.1km/s... barely more than GEO. So, yeah, I reckon a technology demonstrator is definitely doable right now.
Even better --- they've just demonstrated the ability to go to GEO, which is about 14km/s from the Earth's surface. Lunar orbit is only another 2.4km/s, and the moon's surface another 1.6 on top of that!
Chances are that with the technology they have right now, that is, using a modified F9 with the GEO upper stage, they could send a probe on a free-return trajectory to the moon. Or even easier, an impactor. I suspect they won't; Elon Musk appears to have his sights firmly set on the upgraded F9 Heavy and the rocket-landing Dragon, and with that setup you could probably remote land a complete Dragon capsule. I'd be really interested to know what sort of delta vee the Dragon's internal rockets will have...
Luna-9's pictures were sent back using one of the standard encodings used for wireless newspaper photography transmission. During the transmission, the Jodrell Bank radio telescope in the United Kingdom was listening in (well, wouldn't you?) and the astronomers there recognised the encoding, phoned someone at the Daily Express, and as a result the first pictures from the surface of the moon ever were printed in a British newspaper while the USSR was still wondering what to do with them.
There is some speculation that the encoding scheme was picked deliberately to make sure this happened...
The RPi is an ARMv6, while this (along with pretty much every other modern ARM device) is an ARMv7. The ARMv6 has hardfloat but implements a slightly different version of the spec. Most OSes have standardised on the ARMv7 version which means that their code won't run on the ARMv6. So Debian armhf will run on this but will not run on the RPi: you have to use Raspbian instead, which is a version of Debian specifically compiled for the ARMv6. (Of course, Debian armel will run on both, but then you don't get any hardware floating point support.)
The Broadcom GPU is significantly awesome. It is, however, almost totally undocumented. There's a reverse engineering project which has mostly nailed down the instruction set, and there are even some C compilers for it (one of them is mine!) even though there's no gcc or LLVM support for it. You can write programs in C and run them on the bare metal. Unfortunately the GPU doesn't support double-precision float and the MMU is kinda weird, and it's probably going to be slower than the ARM for non-DSP-heavy code anyway, so it's unlikely you'll see Linux for it any time soon. But it's a beautiful, beautiful architecture to write code for. (And it's dual core! Not very many people know that...)
The total amount of power used here would be negligible, of course. But I'm surprised they didn't come down on him for improper disposal of mercury...