You've done some good discovering here, Hakan.
PIE initial *dh- does indeed give Latin f-. There are lots of examples.
Latin initial f can also come from PIE *bh, or *gwh.
And yes, Latin /d/ would normally correspond to English /t/, German /z/, as
you suggest.
You say:
> a newbie question - what exactly is the sound '*dh-'? Is this the same
sound as in English 'there'?
That's an oldie question, too! Disputed, and it doesn't make sense, and it
looks highly unlikely, but we're apparently stuck with it.
(a) The series of sounds represented by PIE *bh *dh *gh *gwh were probably
voiced sounds of some kind (although in some languages, for example both
Latin and Greek, they arrive in the earliest stages as unvoiced sounds).
(b) They were also probably stop consonants (your suggestion of a voiced
fricative such as found in English "there" makes some of the later
developments harder to explain - make dialects show them as stops.)
(c) There were different from the ordinary voiced stop consonants
reconstructed as PIE *d *g etc. (Some PIE dialects show no difference, eg
HIttite, Tocharian, Baltic,Slavic, Avestan, but those that do show a
difference do it consistently).
Various suggestions have been made as to what they actually were.
Interstingly, various and contradictory descriptions have been given of
modern languages such as those found in India today, which contain sounds
conventionally transcribed as voiced aspirates. Some speakers/researchers
insist on a true aspiration following a true voiced stop, while others
describe other phenomena. For PIE the discussion continues.
You say:
>Zeus was called Deus. And we've got the 'dios-kouroi'
This isn't a surprise. The oblique cases of Zeus are formed from the stem
Di-, which turned into Z- in the nom. and acc., so it's suggested it goes
back to a PIE root *dy-. The root is quite complex, with different ablaut
forms, and it sounds as if you've got access to Pokorny, so it's worth
looking up - page 183.
Peter