Mixing Alignments

There is a fair share of possible things to do with alignment. The most obvious bit probably has to do with split-ergativity. However, there are more things we can split, so let us look into these things.

First, we can observe the following basic setups of case-marking for direct objects and subjects:

Nominative:
Subj[1] Obj[2] Verb
Subj[1] Verb
Obj[2] Verb

Ergative:
Subj[1] Obj[2] Verb
Subj[2] Verb
Obj[2] Verb

'Anti-ergative':
Subj[1] Obj[2] Verb
Subj[1] Verb
Obj[1] Verb

Unattested:
Subj[1] Obj[2] Verb
Subj[2] Verb
Obj[1] Verb 

[See source at footnote 1] 

The last one is quite obviously 'weird' and its unnaturalness should be obvious. If the difference between the nominative and the anti-ergative is unclear, look at the "intransitive object" case - it is the same as the subject case in the anti-ergative, but the same as the "transitive object" in the nominative.

What we mean by 'transitive' vs. 'intransitive' objects here is not necessarily obvious - in Finnish - which partially follows the anti-ergative alignment, it basically means that there is no possible nominative subject, which happens when the verb has a quirky case subject, is imperative or passive. The object is still syntactically an object (and pronouns have a distinct accusative marking in this situation), and the differential object marking that Finnish has still applies - negative or atelic verbs take partitive objects.

We can of course create a similar system for ditransitives:

dative:
Subj Obj[1] IObj[2] Verb
Subj Obj[1] Verb
Subj IObj[2] Verb

secundative(?):
Subj Obj[1] IObj[2] Verb
Subj Obj[2] Verb
Subj IObj[2] Verb

antisecundative:
Subj Obj[1] IObj[2] Verb
Subj Obj[1] Verb
Subj Obj[1] Verb

antidative:
Subj Obj[1] IObj[2] Verb
Subj Obj[2] Verb
Subj IObj[1] Verb

Again, the last one seems unnatural, but for some reason less impossible than the unattested antinominative. I have no typological data regarding these, and that is why I affixed (?) to 'secundative'.
Now, we could of course do a simple thing and combine any of the first set and any of the second set and get all kinds of weird shenanigans, where the same case could mark intransitive subject, monotransitive object and ditransitive indirect object, and we could have all kinds of odd combinations about what happens when there's Obj and IObj but no Subj and so on.

However, let's not go down that route right now. Let's instead look at having subsystems that follow different systems among these.

In Baltic Finnic, an ergative subsystem kinda appears with regards to existential subjects - intransitive ergative subjects are marked like (some) objects. However, some "intransitive" objects are marked like some subjects, so an antiergative subsystem is also present, and probably even more present than the ergative-like subsystem.

So, what could be reasonable ways of having subsystems such as these? I dunno, there might be any number of things that could distinguish them, and a language could probably have as many as three or four or five different systems, with quite different lines of demarcation determining which is in use at what point.

[1] Terho Itkonen, Subject and Object Marking in Finnish: An inverted ergative system and an "ideal" ergative sub-system, in ed. Frans Plank: Ergativity, Towards a Theory of Grammatical Relations, 1979, Academic Press