In my opinion there are three main reasons for that.
First thing is that there are two different on-click (or: on-tap) functionalities of these:
- clicking/tapping the text: just shows more details about likes/comments
- clicking/tapping the button: triggers an action for liking/entering a comment
There is a different importance for the user between seeing something and triggering a change (actively liking or leaving a comment). Representing the first with a bare text and the second with a button reflects it quite well.
Secondly, both "Like" and "Comment" have double meanings: one is verbal ("to like", "to comment"), and the other is substantival ("a like", which is a neologism, and "a comment"). Grouping these two meanings would lead to mixing these contexts, so it would not be good for the UX. If you do so, you would have to visually distinguish the two meanings:
- by adding explanatory text for the verbal part, an example: "Likes (43) [Click to like]"
- or by adding explanatory text for the substantival part "[Like] People who like it: 43)".
The other reason may be the tap zones. If you group both verbal and substantival, you would have to make sure it's easy for an user to click/tap the one he wants (due to the different importance). This would result in a necessity of creating quite big click/tap zones for checking the number of likes/comments (if they were left small, there would be a risk of triggering the actual like/comment, inconvenient for the users). This necessity would lead to a significant waste of space, as extending the numbers would eat up space now used for the buttons. Being given a limited width (especially on mobile) it is therefore good to split buttons from numbers, which allows for quite big buttons and quite separated "number of" texts (grouped together, as misclicking/mistapping does not actually trigger any important action from the user).