Regarding the horizontal axis, and within the same high, I'd like to know whether back vowels (e.g. /ɤ/) are more sonorous than front ones (e.g. /e/).
There is some evidence indicating that back vowels are more sonorous than front vowels, although the relation could also be based on round (more sonorous) vs non-round, since the data comes from Modern Greek where back and round correlate in the typical way. This is discussed in E. Kaisse's Harvard dissertation.
There is a general phenomenon of hiatus resolution whereby V#V sequences reduce to a single vowel by deleting one of the two vowels. Which vowel deletes depends on the properties of the vowels; there is also variation between dialects; there are multiple rules that implement this hiatus-resolution plan; and there are different syntactic conditions on the different rules. The general pattern (most pan-Greek) is that the "more sonorous" vowel is retained, and the hierarchy of vowels, in terms of sonority, is [a > o > u > e > i]. Examples of this here come from Peloponnesian (glosses are my construction based on available information from Kaisse).
/ta éxo/ → [táxo] (I have them); /me aγapái/ → [maγapái] (he loves me); /ta onirévome → [tanirévome] (I dream of them); /to alázo/ → [talázo] (I change it); /to urliazi/ → torliázi] (?he howls it); /tu oðiγó/ → [toðiγó] (I lead to him); /mu éðosa/ → [múðosa] (he gave to me)
Interestingly, the system is rather different in Urban Athenian, which has more specific sandhi processes and a different sonority hierarchy, viz [o > a > u > i > e]. There are other claims regarding sonority-based vowel rules in Greek such as assimilation in Cretan which confirms the general pattern though collapsing /e,i/ into a single position. The interesting thing about the Greek pattern is that it seems to contradict the typical functional explanation for the effect, that more sonorous vowels have a more open vocal tract. Round vowels have a more closed vocal tract. If "scale" effects are always based on sonority, then this calls into question the functional explanation.
The other type of oft-cited evidence regarding sonority is that schwa is often treated differently in stress systems, viz. is skipped for stress. This leads some to say that central vowels are particularly non-sonorous, but that does not address back vs. front.