The real drawbacks of using PostgreSQL: your hosting may not support it, the application you want to use may only support MySQL.

In the olden days, MySQL was synonymous with using the MyISAM storage engine, which was not very reliable but was faster than the alternatives (the InnoDB storage engine for MySQL or PostgreSQL.
MySQL+MyISAM was also far easier to setup and configure than MySQL+InnoDB or PostgreSQL.

And the logic was that for most websites, speed was more important than reliability. There are, however, two problems with that logic.
Problem 1: Your website may not need full ACID compliance level reliability, but it needs some. It's mighty inconvenient when then system crashes and boots to a corrupted database. It's like when Windows used FAT32 instead of NTFS.

Problem 2: Nowadays, MyISAM is actually only good at simple, read only, workloads. Complicated queries or writes tend to bog it down. And even simple websites have growing amounts of both.
Nowadays, InnoDB has improved to the point it is, in general, faster than MyISAM while providing better reliability (and it's finally been made the default MySQL storage).
PostgreSQL also improved a lot in performance. In particular, it tends to be better at handling complicated queries and at scaling better in multi-CPU systems.

You're not missing anything, running a bridge at your home is fine.
But since you're not willing to spare your scarce bandwidth, then AWS instance is an easy and cheap way to contribute.

What you suggest is being done, see Sender Policy Framework and DKIM.
And it does help a lot.

However, it's not a solution to end all spam.
- Many domains don't yet publish SPF/DKIM, so they can still be spoofed.
- Spammers often use hijacked legitimate e-mail accounts.
- Spammers sometimes spam from a domain which they do control.

Spamhaus are not self appointed.
We, the system administrators, choose to use or not Spamhaus' black list (or any black list) in our systems to reject potential spammers.
And we, the system administrators, are responsible for consequences of choosing to use a black list in our systems, including the possibility of rejecting legitimate messages and users and all that stems from it, from complains from your users to your boss yelling at you because the e-mail system rejected that important e-mail he was expecting.

Many of us choose to use Spamhaus' black list because they do a good job at a) identifying spammers b) keeping legitimate users out of their black list. They help us keep our jobs.

And if you care, Spamhaus are in a better position to identify what is spam and what is not than most.
They have spent the last 15 years building up the means and experience to identify spammers, without falling victim of such simple tactics as complaining that you're being spammed by a mailing list you've subscribed.

You've got some weird setup there. It has to be doing some state tracking to do NAT or else your outgoing connections won't work. Unless you've set a specific high/low port range in each computer..

Anyway.. whatever rules you have there for IPv4, you can set them for IPv6 as well and it will be less work for your CPU.

Regarding the address space: yes, there are probably many possible avenues.
Never quite stopped to think much about them, I just firewall my IPv6 networks.

Your router already implements stateful IPv4 NAT. Implementing a comparable (or better) IPv6 stateful firewall will put a similar or even smaller load on it's CPU.

Also, you might not need a stateful IPv6 firewall.
The premise is that the minimum IPv6 network your ISPs should assign you is a /64, which has 2^64 possible addresses, which is too large to be scanned.
By using an appropriate address assignment scheme (stateless autoconfig or random DHCPv6), it would be impossible for a potential external attacker to find your devices' addresses via a network scan.
So, unless your device exposes it's address on the Internet in some way it's safe and this should be enough for things like printers, TVs, etc.
Devices like your PC, tablet or smartphone are more likely to expose their IP address (ie, via participating in a BitTorrent swarm) but those need to be able to cope with being on a hostile network anyway. Ie, consider when you use your tablet in a hotel's WiFi.

Then again, this may be just wishful thinking and we'll need IPv6 stateful firewalls!

Like most cable providers, Comcast uses the 10.0.0.0/8 private address range internally to, among other users, manage their clients' modems.
That is, in addition to the client's public IP address, each modem gets a private 10.0.0.0/8 address for management purposes.
Their problem is that they have so many clients.. they ran out of private addresses.
So, they want to deploy IPv6 earlier than most.

Some people just like the concept of open source.

Also, the drivers being closed source, you are at the mercy of NVIDIA for features, fixes, etc.
Now, NVIDIA has generally been good -- they know they have paying customers using their hardware in Linux.
But there are some negative aspects, which NVIDIA has neglected.
NVIDIA proprietary drivers don't implement all the features of the Linux graphic subsystem and so, there are some corner cases that don't work. The most obvious one nowadays is the support for Optimus and the cumbersome way it's being implemented.

Occasionally, it takes a while for NVIDIA to update their drivers to make them compatible with the latest kernel or X.org server.

The kernel code is mostly processor agnostic.
But someone has to deal with the processor specific details and that someone is the kernel. So it also has some very fundamental processor specific bits.
The 386 specific bits just got removed.

It's not a particular problem with hydrogen or even turbine engines. Pretty much all thermal engines benefit of having a cold air intake
Broadly, two reasons:
- Colder air is denser and it takes less effort to feed more air into the engine, in order to be able to burn more fuel.
- Higher temperatures (can) yield higher efficiencies, but the engines are limited by what engine materials can whistand.

Turbocharged petrol or diesel engines usually have a intercooler to cool down the air between the turbocharger and the engine itself. Some gas-turbine electric plants in hot locations pass the air through a room full of ice before intake.

What these guys did was, taking advantage that they use liquid hydrogen (very cold) as fuel, they cool the air down before the intake, making life easier to everything that comes next.
The tricky part really was designing the heat exchanger.

It actually depends on the plant type, not the fuel.

All traditional thermal power plants suffer of the problem, independently of what they use: coal, oil, gas or nuclear.
As Smidge wrote, they use massive steam boilers which take lots of time to heat up and cool down.

Combined cycle power plants are much better, as the gas turbine can be throttled quickly and the steam system is much smaller and hence also reacts quicker. But the gas turbines can't burn coal, unless it goes through a complicated (expensive) process to convert into a synthesis gas.

So, in practice all coal power plants are traditional thermal plants.
There are old thermal power plants burning oil and gas, but most of the new ones are combined cycle --cheaper and more efficient.

AFAIK, oil is getting less and less used to produce electricity as it's just too expensive compared to coal and natural gas.

SR motors are not "far more efficient" than permanent magnet or the other types of electric motors used in high power applications.
All of these motors can achieve efficiencies in the 90% range. The differences come down to few % efficiency, cost, weight and other factors depending on application.

For example, Renault has chosen wound rotor synchronous motors for the Zoe, which is a very rare choice.
But this allows them to re-purpose the motor driver as charging circuit and gave the Zoe a built-in 44 kW charging ability for almost free.

GIMP depends on GTK, not GNOME.

I'm afraid you're incorrect.
I have the Cadence binaries at hand and they're all dynamically linked against a large set of libraries.
What they (and most developers like they do) is to specifically target a reasonable set of distributions, RHEL4 and 5 in particular.
Running on other distributions mostly works, but they provide no guarantees and no support.
In fact, I can't launch 64 bit Virtuoso because of an incompatible Qt library.