verb agreement in one stage of English

Dear all,

I have learned from one book on historical linguistics that the
English perfect aspect(has/have V-en) is derived from a stative,
possessive construction, i.e. from a construction like 'I have a few
ribs broken.' According to the author, there is no indication that I
did the breaking. And the sentence means 'I have a few broken ribs.'

Later, this construction was reanalyzed, or reinterpreted, in a
perfective sense. Hence it is the origin of the present-day English
perfective aspect.

Now, I have encountered a curious sentence as follows:

Ac hie haefdon tha¡Khiora mete genotudne
But they had then¡Ktheir food used-up
'But they had then used up their food.'

The sentence can be only interpreted as a perfect construction
because something, being used up, cannot be possessed. That means
what had been the original main verb(haefdon) was now an auxiliary
and what had been an adjectival function(genotudne) was now the main
verb.

However, something about this sentence deserves our attention: the
suffix is attached to the verb 'genotudn-.' That is, the main verb
now agrees with the object 'mete' only!

Isn't this phenomenon a counterexample against the implicational
universal that states that if a particular phenomenon applies to
direct objects, it should also apply to subjects?
Ths implicational universal also predicts that there are no languages
in which the verb agrees with just the direct object. But in the
above case, the universal doesn't seem correct.

I understand the behaviour demonstrated by the English perfect
construction was at a transitional stage. Therefore it still retained
a residue of the possessive construction.
But is it possible for a language at a transitional stage to go
against the implicational universal? Or the sentence was not evidence
against the universal at all?

I would appreciate your replies.

Ray