Well, this is far more avanced theory of music than you study at university. But if you read any good music theoricians, they will say exactly the same thing as I. Music theory that is teached at school is has been simplified a good deal and is based only on the results of long researches by theoricians of the 16th-19th century about egal temperament. The dream (that now has become a kind of rule-without-wit) of the egal temperament was actually to be able to make fit a broader range of tonalities through a minimum of keys (12 actually)... but any non-keyboard player of any talent will tell you that they never play exactly on the same pitch depending of the notes context... and this famous 'context regulated by the ear of the instrumentalist' has it's own rules which are demonstrated in my charts.
So that's why if you play a G (498) and want a pure major third below you need the Eb... but if you pass after that from G to it's pure major third higher which is B (498+386 (M3)) = 884 and if you wish to modulate in B sharp after that for example.... then you need the major third of B which isn't the Eb but really the D# (for 884+386=1270... to which you got to remove one octave (1200c) to understand more easily and get 70... which is the 71c)... of course... that process would need very far modulations... but Listz and Wagner have done so (and others).
Just to remember so you understand better... a pure third (the only one that will really sound just) has a value of 386c and this is physical... it's a question of harmonic distance in the nature of the sound.
Hope it help... it's not something to be grasped instantly... so don't worry if it's still a bit hard to get !
