Why don't we have 2 Summers and 2 Winters?
Due to Earth's elliptical orbit, its distance from Sun varies by almost 5 million Kilometers (147 million Kilometers at closest point & 152 million Kilometers at farthest point, i.e. almost 3% of the average distance).
As evident from the fact that that Venus has hotter environment than Mars due to their respective distances from the sun.
Why then Earth does not observe two winters (at farthest points) and two summers (at closest points)?
I know that Earth's seasonal climate change is caused by its 23 degrees tilt that causes the sunlight density variations for the hemispheres.
But to me this 5 million Km distance seems more relevant than the 23 degrees tilt.
One problem with the question is that the closest and farthest points only happen once per orbit. See the Wikipedia article: http://en.wikipedia.org/wiki/Elliptical_orbit
"But to me this 5 million Km distance seems more relevant than the 23 degrees tilt." -- It may seem that way to you, but our planet disagrees with you; the tilt has a much stronger effect. (I don't have the time or math to prove it.)
Are you thinking that we'd have one summer at when the northern (for example) hemisphere is tilted toward the Sun, and a second summer at perihelion? One of several problems with that is that the perihelion is in early January, quite close to the northern midwinter.
The sun is at one focus of the elliptical orbit, not at its center. Perihelion is when the earth is at the end of the major axis nearest the sun, not (both times) the earth is at the end of the minor axis. There is only one per year. The 3% change in distance makes a 6% change in solar energy received. At San Francisco, the day is 14:45 long in June and 9:33 in December, an increase of 54% and we haven't accounted for the higher angle of the sun in the sky.
The extreme temperatures of Venus and Mars are only partly explained by their distances from the Sun. Venus is affected by a runaway greenhouse effect; it might be a lot cooler if its atmosphere were thinner. And Mars might be substantially warmer if its atmosphere were thicker (it would probably need to be larger to hold onto a thicker atmosphere).
There are a few incorrect assumptions in your post, so it is difficult to answer as asked. But I can address the misconceptions.
1. The seasons are not caused by our distance from the sun
The seasons are caused by the 23.5° tilt in Earth's axis. When the Northern Hemisphere is tilted towards the sun (summer), the Southern Hemisphere is simultaneously tilted away from the sun (winter). So the seasonal temperature difference has little to do with the Earth's position in its elliptical orbit. Without this tilt, there would be no seasons and the temperature day to day across the globe would be relatively uniform.
2. Even the GLOBAL temperature is NOT consistent with our change in distance
As a matter of fact, the average temperature of the Earth globally is hottest when it is the furthest from the sun — hotter by about 2.3°C (ref). That's because there is a lot more landmass in the Northern Hemisphere facing the sun (when Earth is farthest away in its orbit). So even though there is less intensity of sunlight, the land is able to be heated up much faster than the vast oceans which have to be heated at perihelion.
This distance-temperature inconsistency isn't unique to the Earth. Look at the average temperature of the other inner planets as we move away from the sun:
- Mercury (167°C)
- Venus (460°C) ← farther, but hotter than Mercury?
- Earth (14.0°C)
- Mars (-60°C)
Venus is actually warmer than Mercury because of the thick carbon dioxide atmosphere causing runaway global warming. So it isn't simply the distance from the sun that determines the average temperature of a planet.
3. There's only ONE aphelion/perihelion
The closest point of the Earth's orbit (perihelion) and the farthest (aphelion) only happens once per year; not twice. That is because the elliptical orbit of the Earth is such so the sun is at one of the foci, not the center (as illustrated below).
Note that the size of the bodies and the eccentricity of the orbits are greatly exagerated here.