Here’s my original suggestion - once again, this is my guess, and it can be wrong.
Economy might stand here for the most efficient use of space/materials/other resources at the disposal of the scribe because, as Roger Bagnall puts it,
"the labor involved in making papyrus was considerable, and its price was therefore significant, most typically several ...
Lobel-Page (p. 37) give .[.......] γὰρ ἐφίλει δυ[, saying, "Sub coloph. 2076 schol. vestigia...quod quorsum spectet obscurum est." I'm not trained in papyrology (I much prefer inscriptions—far easier to read!), so it's difficult for me to judge, at least without seeing the papyrus in person.
I wonder if this "principle of economy" concerns the text rather than the physical papyrus itself. That is to say, it is perhaps a principle of textual criticism rather than papyrology.
I say this because that was more M. L. West's field (please see edit below) and also because the excerpts you give seem to draw a distinction between the physical aspects ...
This is actually LP 66(c), from a papyrus published directly in Lobel's Σαπφοῦς μελῶν. From LP's text and notes, we get his transcription as follows:
] . ΚΑΤ̣€Γ[
Where some options for the vestige in l. 2 are ινρ and l. 3 could end in a nu or lambda too. To reconcile my transcription with his:
I see now I missed the top of the gamma in l....
Economy in this context could be the well-known, general principle of economy from linguistics (Passy, Martinet, Tauli etc.) - a possibility which I initially discarded as the most obvious.
E.g. Valter Tauli argued that an ideal language "must contain the maximum possible economy which is compatible with the absolute clarity and necessary expressiveness."
I have to admit, I don't know. I've done some research and I am not entirely sure. So, I offer two explanations - in two different answers.
The principle of (formulaic) economy
As Russo reminds us,
"Anyone who reads Homer in Greek becomes eventually aware that repetition is constantly at play, some of its forms being more immediately evident than ...