Molecular engineering of two-dimensional π-conjugation: expected and unexpected photophysical consequences of a simple particle-in-a-box approach