Assuming that most of us would rather put the old Yale system out to pasture, that leaves two dominant systems of Romanisation for Cantonese: Pinyin (教院式拼音) and Jyutping (粵拼). As they each have their merits, and are both quite similar to each other, I decided that it would perhaps be best to merge the two systems for the sake of unity.
Jyutping concessions to Pinyin:
1) All instances of 'yu' become 'y' (this form is more concise and represents the IPA value of this sound).
2) All instances of 'eo' become 'oe' ('eo' and 'oe' are allophones in character readings*).
3) All instances of 'eoi' become 'oey' ('y' is the more phonetically accurate coda for this final).
Pinyin concessions to Jyutping:
1) All instances of 'dz' become 'z' (this form is more concise, though less phonetically intuitive).
2) All instances of 'ts' become 'c' (this form is more concise, though less phonetically intuitive).
3) All instances of tones 7, 8, and 9 become 1, 3, and 6 respectively (stop-consonant codas render the former group redundant).
*I've been told that 'eo' has since become phonemically distinct from 'oe' in recent years, but exclusively for loanwords and onomatopœia. Perhaps some of you here can chime in on this issue, as it is news to me (my understanding was that only 'aa' and 'a' were phonemically distinct paired vowels.
Jyutping concessions to Pinyin:
1) All instances of 'yu' become 'y' (this form is more concise and represents the IPA value of this sound).
2) All instances of 'eo' become 'oe' ('eo' and 'oe' are allophones in character readings*).
3) All instances of 'eoi' become 'oey' ('y' is the more phonetically accurate coda for this final).
Pinyin concessions to Jyutping:
1) All instances of 'dz' become 'z' (this form is more concise, though less phonetically intuitive).
2) All instances of 'ts' become 'c' (this form is more concise, though less phonetically intuitive).
3) All instances of tones 7, 8, and 9 become 1, 3, and 6 respectively (stop-consonant codas render the former group redundant).
*I've been told that 'eo' has since become phonemically distinct from 'oe' in recent years, but exclusively for loanwords and onomatopœia. Perhaps some of you here can chime in on this issue, as it is news to me (my understanding was that only 'aa' and 'a' were phonemically distinct paired vowels.