You'd be best off turning to a discussion of Byzantine ethnography for a detailed answer, but to put it shortly the Byzantines tended to revive a lot of classical/post classical stuff when it came to ethnonyms. They're pointed anachronisms and the whole practice is rooted in the second sophistic.

The names aren't always consistent, so sometimes the French are Celts (apparently? I can't remember seeing her use that term but it's been years since I've read it), sometimes Galloi, sometimes appropriately Frankoi - but then this term can be applied to Western Europeans generally depending on author.

It shouldn't be seen as an ideological choice. Generally speaking, Western Europe was of little interest to the Byzantines, and their ignorance of the region is reflected in the way they discuss the European peoples. Komnene is a good example of this; another case for this argument is the Byzantine historian Niketas Choniates, who wrote his works at the end of the 12th century. When discussing the crusaders of the 4th crusade, he names them 'Latins' or barbarians, only occasionally referring to their 'tribe' (and then often incorrectly). He described the Marquis of Montferrat as a 'Frankish king'. Generally speaking, in Byzantine eyes there was not much difference between the Western barbaroi, whence they were not named by the names they gave themselves. Other barbarians 'suffered' the same fate. The Cumans and Pechenegs are referred to as 'Scythians' or 'Tauro-Scythians', as were the Avars before them. Older Roman names were still in use in Byzantine times and kept on being reapplied to new peoples on the fringes of the empire.

Simply put, the Byzantines followed the old Roman tradition of naming a people after the country they resided in rather than using the name they used themselves.