I don't think it would be that bad. I know here is this idea that early on people thought shooting off the first nuke down here would ignite the atmosphere, or something--but, as Sagan would say, we are star-stuff. There is another word for that though--stellar fallout. Once a star "burns" down to iron, that's all she wrote. It collapses. That collapse allows one final forging of elements allowing for heavier elements (that and neutron star collisions for gold it seems)
So we are nuclear ashes. Matter is tough. Now if nature allowed all these over-unity devices the conspiracy-believers think are out there--forget shooting off a nuke--I'd be scared to light a match.
In terms of satellites, it looks like they keep getting bigger:
Optics, the size of antennas, all these limit the real utility of small sats, which just clutter up LEO.
There have been even larger concepts: http://www.astronautix.com/craft/globis.htm
The largest was the orbital antenna farm (OAF) called for by Director of COMSAT Labs, Burt Edelson:
Large antennas allow for better reception, but I think we still are addicted to the Smaller faster cheaper mantra of Goldin.
Ironically smaller circuits in some ways led to larger comsats anyway.
Early on, Clarke wanted manned communication platforms, not unlike--I hate the acronym-- OAF,
thatwere large affairs. The idea is that you re-use the dish and solar power systems, but just slide in and out different breadboards.
The Soviets just used off the shelf electronics and made big pressurized, air-conditioned housings for their early systems. The electronics and early nukes were heavy, allowing their space advocates to say--don't shrink the payload--make the rocket bigger.
They still make the Vostok hull to this day, it is just used for research craft like FOTON
Our warheads were small, and our sats were small. So our craft had to be space-rated...rad-hardened..vacuum proof...shock-proof...
Now simple solid state systems, while still better than room sized computers like ENIAC what with vacuum tubes, were rather hardy due to their primitive natures.
Early integrated circuit chips didn't have a lot too them, and only a hadfull of people worked with computers on the ground.
But once microchips were used in numbers down here--a funny thing happened. Computers here evolved faster than what could be put in space. By the time the 486 was space-rated, it was already obsolete. I think it takes eight transistors to hold one bit of data with authority IIRC. Now everyone down here has faster computers than many systems in space.Asia-Pacific uses who don't have fiber-optic trucks or other land-line infrastructure demand more and more data.
So comsats have gotten larger, not smaller. Since Russian rockets were already over-powered, they picked up the slack. Ariane 5 was going to launch the Hermes, and was considered overlarge. Now, it straines a bit and had to be up-dated with the ECA and other life-extensions. Musks rockets are overpowered with regards to EELVs, but that allows him more payloads.
Now some comsats have had to use electric drives since their hydrazine thrusters became stuck, so there is a trend towards all electric propulsion now, and that allows a lower "wet mass" as it were. But there are still adherants to the cubesat movement. I on the other hand think we need to embrace the gigantism, and go farther, with smaller numbers of larger, perhaps shielded systems.
One more note. I rather like primative electronics. To me, the circuits on the viking spacecraft are about what any spacecraft bus needs. Not as sensitive as some of todays circuits, but smaller than Apollos. Some happy medium needs to exist for hardy systems that can last. Voyager seems to be doing quite well after all.