Good point on the "chicliday".
Greek elements in scientific names are common, but the pronunciation of Greek transcribed into Latin follows Latin rules; the Romans wrote Greek as they heard it, and most modern Neo-Latin neologists have followed the Roman conventions. There are exceptions, of course; I can't think of any fish examples off the top of my head, but Palisot de Beauvois named a snake genus "Agkistrodon", which is a direct transliteration of the Greek; subsequent herpetologists tried to emend it to "Ancistrodon", which is how a Roman would have spelled it (and how both Greeks and Romans would have pronounced it).
The pronunciation of Neo-Latin names by local vernacular conventions is a recent and still incomplete change that has only set in since the traditional classical education began dying about a century ago. You can see it in the way many scientists insist on "ch" being pronounced as a hard "c", but let simple consonantal "c" followed by "e" or "i" be pronounced as a soft "c"- just as in our Cichlidae example.
I don't use genuine classical conventions either, I just think the current mishmash of pronunciations is funny.