So it's important to recognize how far down the dialectic we are. This is in response to a version of the #3 defense of philosophy: it may not do anything interesting itself, but it does (at least) produce other things that do. I take it this is the idea – correct me if I'm wrong.
We've admitted this because you haven't objected to the claim that contemporarily, Descartes' geometry or mechanics, and any of Newton's work, are
not considered philosophy. They do not employ primarily philosophical methods; philosophers, in general, lack the competence to read and understand them (unless specially schooled in a particular historical period, or in a particular branch of the philosophy of science); they are not the subjects of textual introductions to philosophy; they are not taught to philosophy undergraduates; contemporary philosophers do not engage with them argumentatively, or attempt to refute or augment them. There is no reasonable criterion on which, then, these works are philosophical in any contemporary sense. Notice that none of the above is true of Descartes' Discourse on Method, or Meditations: philosophers
can understand those, they
do argue with them, they
do write about them in textbooks and teach them to undergrads, and so on.
The reason for this is clear – the
methods employed in the Discourse and Meditations are different from those employed in the geometry. And what's more, these methods
match the methods of prior works of philosophy, all the way back to the Platonic dialogues, and later works, all the way to articles in the journals today. So it is clear at present that there is a continuity between
these works, precisely the ones you are not willing to defend as interesting natural science, and philosophy, but it is
not clear that there is any interesting continuity between those works you are willing to defend as interesting natural science and philosophy.
So you must be making some weaker claim – they are not philosophy as contemporarily understood, but maybe at one time they were thought to be? Or maybe even though they're not philosophy in any sense, at least they resemble philosophy in some interesting way? Or what I think you are likely saying, and which is really what #3 is getting at: historically, they developed
out of philosophy in some interesting way, though they're distinct.
Unfortunately, this is a much harder historical claim to prove. It is obvious that works of philosophy develop out of each other, because they cite each other, and their sole reason for existing is the fact that the author read some other works of philosophy. This is not at all obvious for Newton's work, because, as you admit, most of it has little to do with what philosophers were doing at the time, or even had done in the past. Are a couple principles appealing to notions like 'cause' enough to establish a historical connection? I would hope not – or else philosophy 'wins' by default, since it talks about everything, so everything is philosophy (#2).
So let's ask a more productive question –
what led Newton to write the Principia? What methods did he employ in framing the principles he did? Was reading philosophers the primary motive behind this?
Would the work have been writable in the absence of those philosophers? Are his goals or results philosophical in any interesting sense, by either contemporary standards or 17th c. standards? And no, it's not enough to say 'ah, but Newton had so many philosophical implications!' etc. This is because since philosophers can talk about
anything, this move can be used to trivially claim that anything is philosophically relevant and therefore philosophy (#2).
We also need to ask why, even if somehow we thought this was 'philosophy turning into something else,' philosophy's historical core always defaults to what's present in the Platonic dialogues, and nothing else.