Apples n' oranges. Re: your window tricks: all that results in air flow past un-aero surfaces resulting in high drag. The headlight bucket with almost no flow-through (just a bit around the gap around the sealed beam) is different. I don't think Popular Mechanics would simply regurgitate marketing tripe but I suppose it's possible.
I said leave the windows CLOSED, no trick except to try and describe a larger air bucket.
I suppose parachutes are really streamlined once they fill up with air too
In some cases, having a pocket of air form can actually reduce drag. Look at pickup trucks and the effect of having the tailgait closed versus open.
I doubt this would apply to the Metro's headlamps, but the only way to know with 100% certainty is to test it in a wind tunnel. Testing it for fuel economy in the real world, you may get results that aren't statistically significant.
Since we don't all have access to wind tunnels, those here follow a general rule of thumb that won't yield an optimized design, but will give large and measurable improvement in drag if all of these modifications are taken cumulatively. Basically, this rule of thumb is to make the car as smooth as possible with as few gaps, projections, and holes as possible, and have the rear end be as close to a 6:1 fineness ratio as possible. This may not give the best design, but in cases like Basjoos' Civic, Phil Knox's T100, Darin's Metro, and other vehicles, the results are huge. Some of these aero modifications may have actually added drag, but the overwhelming trend is that they have reduced it.
From what I've read, there are two main methods to optimize a car for aerodynamic efficiency if you have access to a wind tunnel.
These are detail optimization and shape development starting with low drag shapes.
Detail optimization involves starting with a concept based entirely on style without yet having any regard to function. From there, areas such as spoilers, mirrors, grilles, windshield, bumpers are optimized piece by piece to reduce drag, reduce lift, aid cooling, or other functions. You basically start out with a bad shape and end up with something better than you started with that remains true to the basic stylistic design and is very hard, if not impossible, to distinguish from the original stylistic concept with the naked eye.
Shape development starting with low drag shapes involves starting with a basic concept shape, as opposed to the concept rendering of a car. The shapes include the Jaray body, Schlor, Kamm back, Torpedo, half-body airfoil, Reid body, and others. In contrast to starting with a fully stylized car with high drag, these basic shapes don't include any stylistic or functional details. Just basic low drag shapes. From there, these shapes are built into a car piece by piece with optimizing every last detail.
The folks at gassavers have no access to a wind tunnel. They can't custom build spoilers and parts with a garuntee that this shape will produce the lowest drag possible for their application. They are stuck with a car that's already been built and optimized according to the sub-par specs the auto companies decided was good enough to sell. So they don't do either method. Instead, they start with a crappy shape and try to turn it into a streamlined shape. Overall, it works. Even if it may not yield the lowest possible drag given that some modifications won't be optimized or may increase drag, the general trend is very clear. It's sort of like a reverse shape development, only you start with a high drag shape and move toward a low drag one. I guess we could call this the hobbyist method. It's not the best, it's not very fancy, but it works.
Great points Toecutter, wind tunnel access would be our holy grail for aero mods, but it is expensive.
Here are the pointers given in the "Wind Camp" article from the Mar 2007 issue of "Hot Rod", where they actually do take a car to a wind tunnel to better optimize it for Bonneville racing.
5 easy mods that almost always work:
-Lowering the ride
at least 20 counts of drag per inch
-Grille block (yes, in a performance mag!)
15-30 counts of drag and 50-100 counts of lift
-Front air dam (big one - from the front bumper down to the ground)
20 counts less drag, 50 less lift. more effect the taller the car is
-Seal the back of the cowl (stop airflow going under the hood at the rear)
10-20 counts of drag, 50-75 counts of lift
-Remove the outside mirrors
10-20 counts less drag.
*Except in '94-up Chevy Caprices where the Cd actually increases without the mirrors
One "count" of drag is .001 Cd, but they add up quickly
Also, on their list of "aero stuff that really doesn't matter":
smooth wax shows no difference to a bumpy spray-can job in the wind tunnel
works for small spinning spheres. Cars need other ways to delay flow seperation
They tested it on a '70's Camaro and it made no difference.
-Smooth rivets and hood pins
Apparently Howard Hughes was wrong.
-Dropping the tailgate
But the wind tunnel guys did say that extended cabs and crew cabs are more aero
-Windshield rake (after a point)
Once it is past 45*, the wind tunnel guys said don't bother. The mag tested it on their Camaro (the '70's models are around 45* rake) and it made no difference.
BTW: if you want to wind test your own car, go to www.a2wt.com The price is $345/hr for the first two hours and $490/hr after that. But in probably a days testing, they were able to take their Camaro from a .497 Cd to 0.201
I'll also add one more of my own - fill the gap between the bottom of the rear bumper and the trunk well (or fuel tank if it's a RWD). My car had dirt and gravel up in there. If small rocks get trapped, air certainly will. If nothing else, it will keep that area clean and less prone to rust.
I'm just about there myself peakster. I read many ideas about it on teamswift.net and even picked up a couple cutoff wheels for my angle grinder just yesterday for this purpose.
I'm going to go with cutting down the stock coils since that will accomplish the result and several people have taken that route. I've got a spring compressor and will try and do it on the car (without taking everything apart). How many coils, and from which end? I havent sorted that out yet, sorry. I need to take a closer look at things first.
^^You should start a new thread with this question because the answer will probably take a few replys. It is possible to cut the springs in place but you have to be very careful. you will not want to try to use a spring compressor on the car with the strut on the car as this will be a MAJOR pain. As long as you keep it minimal(1-1 1/2") you can get away with it on stock struts but as soon as you get lower than that you need a firmer shock to control the suspension otherwise you will get a bouncy ride. Start with cutting one coil and see how it looks from there, then if there is not enough travel trim the bumpstop a little. I usually cut from the bottom of the spring but it depends on what the bottom of the spring looks like. If it does not taper much in order to sit flat on the perch, you can cut from that end. Post a picture in a new post of your struts/springs front and back and your questions can be better answered.
This subject needs a new thread to fully discuss so this is all I will put here.
For me, at least, using spring compressors on the car is almost impossible. I gave it about 5 seconds worth of try once and then just pulled the strut and did it seperately. I assume the geo has a similar set up to mine...