the wrinkles in the foil will create hot-spots and burn your plant. white paint is better.
This is absolutely incorrect.
Get rid of the aluminum foil
Sigh.
yes the wrinkles will mess things up, but im growing with foil right now, and as long as it doesn't have wrinkles, your fine. it might be worth it just to sit down and take some time to put up some cheap tin foil. then again, Mylar you can never go wrong with.
Wrinkles decrease the surface reflectivity, and increase surface diffusion. This lowers the likelihood of hot spots even when curving the surface to create intentional hot spots.
you can go wrong with mylar just as easily as you can with foil. all the same problems occur.
Mylar doesn't really retain shape like foil. Mylar is a better intense(less diffuse) reflector of light compared to foil or paint(somewhat significantly). Mylar is also flammable, foil is not.
So, what if my walls are an off white? does it need to be a bright white?
You want white paint with titanium as the pigment. Often called flat white, check the label, you want majority titanium dioxide(TiO2), pigmented paint for best results. Titanium paint has a significant specular component, and is roughly equal to regular aluminum foil in terms of PAR reflectance. However, humid environments and paint don't mix, you should repaint yearly to maintain peak reflectivity.
More here:
http://www.wisegeek.com/what-is-titanium-white.htm
Hot spots occur when the surface is parabolic or similarly curved, which can focus light as a lens does. Smaller wrinkles, dents, etc, do not have a significant enough area to focus light into a hot spot. And completely work against intense reflectivity. Indeed, the opposite of 'what's claimed' occurs, actually, and more light is diffusely scattered in 'random' directions.
If you want to test reflectivity of objects, I suggest getting a blue or red laser. Shining it on an object and observing how much of the light is observable from various angles will show you how diffuse or specular the surface is. For example, it's fairly difficult to see a laser's striking point(normal) on a mirror when the laser is directed away from you. It's really easy to do on white paint. it's also hard to see on a black surface, as most of the light is being absorbed.
Indeed, with a simple laser and some experimentation you can roughly determine the relative absorption, reflectivity/reflectance, and whether a surface tends to be specular or diffuse for any surface you have by just using your eyes as measurement.
A good test to determine just how specular a surface may be is to use it as a mirror and observe the reflected image of the laser. The tighter the resulting reflected beam is, the more specular(mirror-like), if the beam turns into a uniform glow, it's highly diffuse(typical white paint, completely diffuse is also called Lambertian).
Now a completely Lambertian surface reflects 1 unit of light over half a hemisphere. In steradians(sr, units of solid angle) 1 sr will become 2*pi sr, resulting in significantly less intense light. It's also significantly more spread out.
More about specular/diffuse reflectivity:
http://las.perkinelmer.com/content/ApplicationNotes/APP_ReflectanceMeasurementsofMaterialsUsedintheSolarIndustry.pdf
http://en.wikipedia.org/wiki/Diffuse_reflection
http://www.glenbrook.k12.il.us/gbssci/phys/class/refln/u13l1d.html