Tarkus,
Here is a simple graphical representation of the relationship between sizes and observer distance.
The grey circle M is 1/4 the size of the green circle E. The blue square S1 is located at a distance at which M appears the same size as E (indicated by the red lines). S2 is located further, at which distance it is clear that E appears larger than M.
As M gets further away, the distance of S1 from M at which the two circles appear the same size increases, but crucially it always remains less than the distance between E and M, and in fact the ratio of distances remains fixed.
This does not change as the distances get larger! Any observer further away than S1 in any of these examples will see a smaller M in front of a larger E. Exactly as in the NASA images, in fact.
Su dibujo es simplista y erróneo: la Luna se encuentra a unas 30 veces el diámetro de la Tierra, mientras que en el dibujo, llega a estar sólo a 10 veces... y notará usted que cuanto más se aleja la Luna de la Tierra en ese dibujo, mayor porción de la Tierra es cubierta por la Luna, por esa razón la Luna jamás podía verse de menor tamaño cuando está en primer plano.
Si usted es astronauta de Apolo y ve a la Tierra desde la cara conocida, se sube a un rover imaginario capaz de llevarle hasta las antípodas de la Luna, una vez usted llega al punto opuesto del alunizaje (en la cara oculta) se sube a un cohete y despega en línea recta... nunca podrá observar la Tierra, pues ésta quedará por siempre cubierta por la Luna. Lo mismo sucede con la sonda automática, un tránsito real de la Luna delante de la Tierra necesariamente debe mostrar a la Luna MUCHO MÁS GRANDE que la Tierra.
Se han registrado algunos tránsitos de la Luna con planetas como Júpiter, usted no puede hacer el truco de que júpiter se vea de mayor tamaño, a pesar de que (obviamente) Júpiter es MUCHO MAYOR que la Luna en tamaño, lo mismo aplica para el caso Tierra-Luna.
I speak spanish.
For anyone wondering, Ill traduce what he said the best I can:
Your drawing is simplistic and wrong: the moon is at a distance 30 times the diameter of the earth, while in the drawing, is just 10 times... and you will notice that the farther away the moon is from the earth in this drawing, a bigger proportion of the earth is covered by the moon, because of that reason the moon would never be seen of less size when it is on the first plane.
If you are an Apollo astronaut and you see the earth from the known face, you drive the imaginary rover capable of carry you on to the farside of the moon. Once you are in the opposite side of the moon from where you landed, you get inside a rocket and you fly away in a straight line... you would never see the earth, because it will be covered for ever by the moon. The same happens with the satellite, a real transit of the moon infront of the earth must necessarily show the moon much more bigger than the earth.
It has been registered some transits of the moon with planets like jupiter, you cannot make jupiter be seen with a bigger size when jupiter is obviously much bigger than the moon, the same example applies in the Earth-Moon scenario.