My impressions...
Tire Pressure -

This should be well beaten into everyones skull by now...
Wind Resistance -
As far as I am concerned, the jury is still out on this one. I remember seeing data somewhere that implied keeping the tailgate up was better for MPG than dropping it or running the mesh gates...
Change Spark Plugs -
Again...

This should be a no brainer to most of us. From what I have seen (SAE papers, Joel5.0, etc.) Standard Autolite plugs will perform the best.
Intake Muffler Restriction -
I removed this 'muffler' on my '87 LX and didn't see ANY change in MPG.
Free Flow Air Filter -
Again... I didn't see any change in MPG on my '87 LX
Disable EGR -
In most ecu cases this guys comments are pure BS. The EGR is typically disabled at WOT, so the impact on performance is non-existent.
In Theory... EGR can improve MPG at part throttle cruise as well, by allowing less fuel to be used, and allowing for more ignition timing, and allowing the throttle blade to be 'more open' for a given vehicle speed.
This hasn't seemed to prove true for most of the guys I have seen posting results online though.
Ram Air -
From what I have seen, it would take 70-80mph at a minimum for 'ram air' to have an impact. Some sources say as high as 100mph...
Throttle Body Spacer -
Beats me what this would do

Synthetic Oil -
Depends on the viscosity being used, but from what I can tell a thinner oil will allow for better MPG.
Fuel Octane -
This would result in higher power if the Ignition Timing were changed to accommodate the better fuel...
Coolant Temp -
I didn't read all of this (it is late), but from what I gather, better performance and economy can be found with lower temps, as long as the engine oil is at a good temp...
EFI Modifications -
I have been working on increasing the air flow through the engine and consequently, I should increase the fuel flow appropriately.
The size of the fuel injector and the mass air flow sensor should be matched together within the limits of the computer's (MCU) programming. If you replace any of the components and they are out of the limits of the MCU's programming, you will end up with a poorly performing engine.
Yes and No....
The ecu can correct for an error in the AFR of about +/- 12.5%
Correcting for larger injectors by running an aftermarket 'calibrated' MAF meter is generally the lesser of 2 evils.
The BEST solution to increasing fuel needs is via a custom tune on a J3 chip.
Fuel Pressure -
Again, the factory ecu is likely to have a range of +/- 12.5%
The ecu will always try to push the AFR toward stoich, as long as it is within the 12.5% range
Trying to tune for power by pushing the AFR away from stoich is a waste of time, as eventually the ecu will drive the AFR back toward stoich.
MAF Sensor -
If you are going to change the stock MAF for another stock MAF, then a J3 chip is the ONLY option as far as I am concerned.
Advancing Cam Timing -
I have yet to play with this on my 2.3na, but I can't wait to do so.
Selecting a 'Hotter' Cam -
I look at it this way... If you have a stock intake/exhaust/drivetrain... and a stock cam.... then you basically have a CUSTOM cam for your application.
Now if you want to do something other than what the factory intended that engine to do, it may be beneficial to run a different cam

Gear Ratio -
Again, :duh:
Exhaust -
I've studied swapping in headers and free flow exhausts for about 40 years now and have concluded that any performance gains you will see from headers will only been seen above 5000 rpm. Around 3000 rpm, you won't seen any performance increase at all.
I'm more interested in REAL experiences...
With that said,
I have a header from a 94 Ranger almost ready to install on my n/a 2.3....
I suspect I will see gains above and below 3k rpm, but a cam timing gear will be needed, as well as custom J3 port tuning.
Intake Modifications -
:duh:

Head Work -

Interesting stuff....