Zevele - I've been laughing everytime I load this page and see "more coherent way".
KingSparta: 1|PLS|1=3 for sufficiently large values of 1.
(actually, 1|PLS|1 is whatever you feel like, but that's really off topic).
Agonostis - some defs (to make sure I know what I'm talking about). A rhombus is 4 sides, and each side is parallel to exactly one other, and all sides have the same length. A parallelogram has the parallel requirement, but does not require all 4 sides to be the same length, right? (doesn't matter too much - your counterexample relies on different angles instead of different length sides, I think).
A "really really skinny" rhombus has two angles of about 1 degree, and two angles of about 179 degrees. It looks kind of like a toothpick. If you position it perfectly, it'll cover a hole, but when you pick it up and drop it, it'll probably fall through the hole, so it's not a counterexample. What am I missing? The "mininum diameter" of the rhombus is the distance between the two vertices of the 179 degree angles, and this is the "maximum diameter" of the largest hole the rhombus can cover, which is a lot smaller than the largest hole a square of the same area could cover (without falling in).
Rest of the world: google on manhole and feynmann and you'll find the best answer.
j