The digital ecosystem as a whole is built on classification and categorization.
Maybe Geoffrey Bowker and Susan Leigh Star ("Sorting things out") would be a very good addition to that. You could also cite Foucault on who Bowker and Star are building upon.
Data Feminist approaches hold significant potential for advancing more ethical research within the Digital Humanities; however, they have not yet been fully integrated in the German-speaking Digital Humanities community
"hold significant potential" is first of all a statement but at the beginning an in the end it seems more like you would like to state a claim. There might be a danger of a "is-ought problem" (D. Hume) or "naturalistic fallacy" (G. E. Moore) so you could tackle this to make the basis of your argument (and the claim if it should be one) even stronger.
gaps
I think this is a very crucial point: "Gaps" can only be gaps if you think from the perspective of a "whole" - but what is this "whole" in this context? This shouldn't be thought of as whataboutism or "if you cannot have it complete leave it as it is" - the problem is the construction of the "whole" itself, especially if you think about intersectionality which multiplies neglected and oppressed groups. The construction of the whole may produce new neglected groups which shouldn't be a side effect.
Second, the history of how those sources became data must be examined.
This is also true for digital-born objects like posts on social media.
political
Altough not explained I think I understand this point and I agree. I would add that that good data (= without bias) is also a matter of epistemic and scientific quality, though.
these
Why this difference between the abstracts?