The gemara in menachos implies that not only is it unnecessary to draw lines in tefillin to keep the writing straighter, but there is an assumption that this was specifically NOT done. Why? Tosafos here and in menachos writes that since you are exempt, it would qualify as a middah of hedyotos to be machmir. Tosafos in Sotah 17b (cited by R. Akiva Eiger) claims that it is actually a halacha l'moshe misinai not to do it. Nonetheless, Tosafos says that if one cannot keep the writing straight without lines, they can draw lines for the say of beauty of the mitzvah. Although our Tosafos implies that there would not be any distinction beween sefer torah and tefillin in regard to beautifying it with straight lines (which apparently is what we rely on to draw lines in tefillin), Tosafos in Menachos quotes Rabbeinu Tam that only by sefer torah is there an inyan of beauty, but tefillin which are covered have no advantage of beauty with the parshiyos.
The Beis Haleivi (end of sefer al hatorah) points out from Tosafos question, "why don't we say that the tefillin had sirtut?", we can deduce that mezuza which does need sirtut does not need sirtut lishma. The lomdus is that the gemara implies that if mezuza needs sirtut, you would not be able to turn tefillin into mezuza, because tefillin is missing sirtut. On that Tosafos asks that the gemara should answer that we may be speaking of tefillin that has sirtut, and only tefillin which had sirtut could possibly be turned into a mezuza (to which tosafos says, it must be that one would never find such an animal, since it is a hedyotos to do that). Why doesn't Tosafos answer that mezuza needs sirtut li'shma, therefore even if one were to do sirtut on tefillin for the purpose of tefillins, since tefillin technically don't need sirtut, it would not qualify as sirtut li'shma, and that is why one would never be able to turn tefillin even with sirtut into a mezuza. From the fact that Tosafos doesn't say this, implies that mezuza would not need sirtut li'shma and therefore if one would find tefillin with sirtut it could be turned into a mezuza (so Tosafos has to say that one would never find such a thing). This is against R. Akiva Eiger in Hilchos mezuza who says that sirtut has to be li'shma.
An entirely separate point - pashut he'ara. Tosafos says that the pilegesh b'giva didn't really commit adultery, because if she had her husband would never take her back. Although, it is technically mutar for a husband to take back a pilegesh after adultery, "it was not the derech to do so". Tosafos proves this from Dovid Hamelech who didn't take back his pilagshim after they were with Avshalom. The meforshim in the back ask that one cannot cite a proof from Dovid Hamelech since our case is different for 2 reasons: 1. They didn't commit adultery with Avshalom, rather they were raped. 2. Dovid Hamelech was different because his wives were "profaned" and therefore unfit for a king (similar to the idea in Avoda Zara 11a that we burn the keilim of a king so that it is not used by a hedyot - just as something used by a king cannot be used by a hedyot, these woman used by a hedyot are unfit for a king). I would suggest that the 2 distinctions compliment one another to explain the proof. If for a normal person it would be "the derech" to take back a pilegesh after adultery, then it should not be considered a bizayon for a king to take her back after being raped because a king would only have to be one notch above a regular person. From the fact that it was a bizayon for Dovid to take them back even after being raped, it must be that for a regular person it is not the derech to take them back after commiting adultery.