What if Nubian cliental kingdoms formed during the Imperial Roman Age ( allegedly there maybe was even one and Nero had plans to regarding Nubia) and Latinize to some degree ?
Romanization ≠ Latinization. In almost all eastern Romania, it took the mantle of hellenistic network of cities (whom laws and customs were integrated to Roman law, rather than coming from colonies of Latin and Roman law directly) and political relationship that was already present and dynamic before the Roman conquest, from which Roman institutions and frames imposed themselves.
Now, to the matter at hand : clentelization was a practice that tended to disappear with the classical Roman Empire, emperors preferring to resort to direct control. It wasn't systematical (especially on the Persian border) or irremediable (Crimea being provincialized then re-clientelized), but the "golden age" of clientelization was passed.
The problem of Roman-Nubian relations is that Nubia was really peripherical (especially as Kushites moved their chief center South) and heavily depended from the situation in Egypt to thrive politically : as Roman Egypt was really recentered on Mediterranean relations since the hellenistic period and even more so with Roman influence and conquest, Kush declined slowly at the benefit of Axum and Somali city-states.
As Rome campaigned against Kush in the Ist century AD and ruined Meroe and the north of the Kushite kingdom, it more or less created a vaacum that didn't broke immediatly their relations with the nubian state, but allowed peoples as Blemmyes and Nobatians to take place in northern Nubia, as sort of blur march of conflicting loyalties.
It's quite interesting to see that Blemmyes (living on the region set roughly between Nile and the sea) and Nobatians (probably issued from the desertified regions of north-western Sudan) did enjoyed more of a clientelized/laetic relationship with Rome than Kush on this regard.
Simply said, Rome was uninterested on Nubia as long as Nubian and quasi-Nubian peoples didn't raided their borders, and historically, you
did have nubian peoples being clientelized or enjoying a relationship with Rome close to what existed with Berbers in Roman Africa.
But it wouldn't really happen with Kush for the aformentioned reasons : we'd rather sooner see a clientelized Axum or Hejaz.
A detail, but Federates ≠ Clients. the federate status as it appeared in the IVth century was more or less a deviation of dedicitii status, as in a personal status of subservience of non-citizens (which was eventually taken as a relation between the emperor and king/kinglets inside Romania) and clientelization, which is is more a deviation of a patronage and sponsoring between two freemen (if of different statutes) closer to a status of protectorate.