Returning to the Netzi'v's distinction between limud=accumulation of knowlege vs. asiya/la'asos=innovation, chiddush, and his suggestion that chiddush, asiya, is acceptable only when done lishma, Anonymous suggested in a comment that this fits nicely with the Chasam Sofer's tshuvah (Shu"T O.C. 208) that the heter to write Torah sheba'al peh, based on "eis la'asos laHashem", applies only when chiddushim are written lishma -- once again, the term "asiya" demands purety of motive. The C.S.'s assumes that the issur of writing Torah sheba'al peh remains in effect even to this very day and is suspended only on an as needed case by case basis provided the intent is l'shem Shamayim. (In addition to Anonymous' other comments to that post see R' Shternbruch's intro. to Moadim u'Zmanin where he grapples with this C.S. and the challenge it poses to anyone who dares to publish; see also Yechaveh Da'as III:74).
In light of the Chasam Sofer's claim that writing torah sheba'al peh constitutes an issur d'oraysa (see Tos. Yeshanim, Yoma 70) it is striking that the Rambam does not once even mention this halacha. Rav Solovetichik explained (Perach Mateh Aharon p. 48-49) that the Rambam did not in fact leave this halacha out. The Rambam understood the prohibition of writing torah sheba'al peh is not an independent issur, but is a function of the fact that certain elements of Torah are categorically designed to be transmitted by mesorah from teacher to student just as other elements of Torah are categorically designed to be conveyed as text. When the Rambam in his introduction to the Yad records the passing of tradition from generation to generation from Moshe Rabeinu to the days of Rav Ashi, that chain of mesorah represents the fulfillment of this halacha which prohibited writing torah sheba'al peh. However, once that chain was broken, once the transmission of mesorah from teacher to student was lost and replacedby written text, this halacha ceases to have any practical bearing. We no longer have an oral mesorah.
Based on this approach, there is no basis for the Chasam Sofer's claim. The reason we are permitted to write divrei Torah is not based on eis la'asos, but is based on the fact that our methodology of transmitting the mesorah has changed from the person to person link that was operative until the completion of the Talmud. This approach also resolves the question raised by R' Shternbruch (and quoted by R' Ahron Soloveitchik in the name of R' Elchanan Wasserman) whether once Mashiach arrives we will still be permitted to use text to study Torah sheba'al peh -- since mala'ah ha'aretz de'ah and there is no chance for forgetfullness, what might be the eis la'asos permissability of using a text? The answer is that it is not the necessity of avoiding forgetfullness which is the basis of our heter to write torah, but rather it is the fact that the knowledge we record was never part of a chain of mesorah that was exclusively orally transmitted person to person. That fact will not change even with the coming of Moshiach.