>In Satem languages *kWa *kWo > *ka *ko and *kWe *kWi > *c^e *c^i
Not really. It is better to think Satem collapses *kw and *k, but maintains
the distinction from *k'. this allows us to distinguish:
(a) later palatalisations from the Satem phonetic change.
(b) the palatalised outcome of original *k' and original *kw/*k
It also allows us to see these later palatalisations as something that some
Satem and Centum languages might do, and some might not - i.e. a different
phenomenon from the satem change, and independent in various languages, in a
way that the satem change is not.
Peter