compactness of consonants

This question was first posted on phoNet, but as it seems there
nothing happens on this list, I re-post here. Sorry for possible OT.

1. Some of my sources state palatals and velars are compact
consonants. Others - that while labials and dentals (and
postalveolars?) are diffuse, the compactness/diffuseness opposition
is just not applicable to palatals and velars. Where is the truth?

2. One of my sources implicitly states that plosives can be
classified in terms of compactness/diffuseness (and
graveness/acuteness, respectively). But the spectrograms I've seen
show the following: for, eg, [bV], [dV] and [gV], we have f1 in the
form
_____
/
, and f2 in the form \_____ for, eg, [ba], _ ___ for [da] and

_____
/

for [ga], on the one hand, and
_____
/ for [bi], _ _____ for [di] and \____ for [gi] on the other,
horizontal line representing f2 for the vowel. As I can see, the
frequencies of the first two formants of plosives depend on ones of
syllabic component, being more close ('compact') or distant
('diffuse') depending on the appropriate characteristic of the vowel
segment. It's the fequency transition as such and it's direction
(different for different syllabic segments!) that makes the
difference. I am confused. Would anybody help me?

Sergei