I will try to provide some helpful information that is gleaned from a combination of personal experience gained in my past career working with an automotive engineering firm, building multiple 800whp cars and personal high performance car. Along with a massive amount of time spent researching online and in books to satisfy my automotive obsession.
Anyone who tells you that they are snake-oil or don't make a difference is way over simplifying and in 90% of the cases wrong. They do make a difference and they can add power. Every engine and vehicle will respond differently and the design of the intake will make a difference. This is not rocket science, an engine is an air pump plain and simple. Fuel is never the problem, adding more fuel is easy, it is getting more air into the engine that is the key to power. Every engine responds differently to different air/fuel ratios and different ignition timing but simply speaking adding air means additional fuel, means increased mixture density in the cylinder, means more power. The poster (catrains) above is right that the ECU is targeting a specific A/F ratio, he is not quite right that it is usually 14.7:1, that 14.7:1 is what is known as stoichiometric and is the ratio of air to gasoline that fully burns the fuel in the mixture. That ratio is rarely the ratio that produces the most power under full throttle, if you were to dyno a 4Runner I would bet that Toyota has it running somewhere between 13 and 14. Cars run A/F ratios all over the map under difference scenarios to produce the best combination of power, economy and emissions. Depending on the car you have it is most likely using a mass airflow sensor (MAF) to measure the "mass" of the air moving into the engine or a manifold pressure (MAP) sensor and air intake temperature sensor (AIT) to measure the air density and volume moving into the engine to determine how much fuel to add.
So, if we come back to the air pump concept we can think our way through this. The engine itself (no intake, no exhaust) has a theoretical maximum amount of air that it can flow or pump. Now, put that engine in a car, add an exhaust with a catalytic converter, long sections of piping that bends and winds, mufflers to control noise, an intake that has been tuned to filter out unwanted noise with baffles and resonation chambers, a filter to catch dust and provide a specified service interval, corrugated tubing to flex and cut down on costs. Well now you have taken that engine and reduced the amount of air that it can pump. Take that car and run it on a 110 deg. day and you also reduce its efficiency because hotter air is less dense than cold air. Take it to the top of a mountain and you have done the same thing but to a much greater extent.
Now we see how improving the efficiency or reducing the restriction in any of these components can increase power. With the "cold air intake" you are increasing the efficiency of the air intake tract. This is done typically with a smoother intake pipe that removes the resonation chambers and corrugation and a higher flow air filter that reduces the restriction provided by the OEM filter. Now, the once caveat here comes from the air temperature and density issue that I mentioned before. If the factory intake system is taking its air from a duct or area that is providing fresh, cool air to the engine and you replace that with a cone filter that sits in the hot cabin air you may have installed a less restrictive system but are now hindering its effectiveness by removing the cool air component. For the 5th gen 4Runners the TRD and AFE intake systems draw the same cool air as the factory airbox but have higher flowing filters and ducting and will increase power. It may not be 25hp like you can see on some turbocharged vehicles but you will pick up some gains, especially under full throttle. The removal of the resonation chambers will provide increased induction noise which is also desirable for many people.
The throttle spacer lengthens the intake tract after the throttle which can, sometimes, provide increased low end torque. The results from these are usually suspect at best, save your money.
