This root first appeared as ᴹ√SED “rest” in The Etymologies of the 1930s with derivatives like ᴹQ. sére “rest, repose, peace” and N. sîdh “peace”, along with the name of the Valië ᴹQ. Este < ᴹ✶ezdē (Ety/SED). The root √SED “rest” along with the etymology of Este reappeared in both the Outline of Phonetic Development (OP1) from the 1930s (PE19/45) and the Outline of Phonology (OP2) from the early 1950s (PE19/92), and again in revisions to the OP2 page annotated in green ink and thus probably from 1970 (PE19/91 and note #110).
The root √SED “rest” was mentioned a couple times in the Quenya Verbal System of the 1940s as the basis for the verb ᴹQ. ser- “rest, repose” (PE22/102, 125) and again (unglossed) in the Quendi and Eldar essay of 1959-60 as the basis for Este (WJ/403). In the sense “peace” this root had some competition, however. In the first version of the Gloria in Excelsis Deo prayer from the mid-1960s, Tolkien used Q. sívë for “peace”, revised to Q. sérë in the second version and then Q. rainë (VT44/32). The word Q. sívë “peace” reappeared in unpublished etymological notes from 1969 derived from the root √SIB “rest, quiet” (VT44/35).
Neo-Eldarin: For purpose of Neo-Eldarin, I would use √SED for the act of resting and so Q. sérë “peace” = “restfulness (personal peace)”, and would use √SIB = “quiet, ✱stillness, absence of activity” and so Q. sívë “peace” = “absence of activity or violence (environmental peace)”. I would use Q. rainë as “good will” in connection to √RAY “smile”.
Tolkien experimented with a variety of roots for “desire, desirable”, some in connection to the name Idril. Perhaps the earliest iteration of this root is indicated by the Gnomish word G. irn “desired, wished for” (GL/52), but given its other related forms such as G. îr- “will, intend to”, G. irm “wish, intention, resolve” and ᴱQ irya “wish” (GL/52; PE13/116), the (unattested) early root ✱ᴱ√IRI probably meant “wish, intend”.
In The Etymologies of the 1930s there was a root ᴹ√ĪR “desirable, beautiful”, but it had no derivatives (EtyAC/ĪR). Elsewhere in the document ᴹQ. íre “desire” was derived from the root ᴹ√ID (Ety/ID), perhaps a new iteration of the (hypothetical) early root ✱ᴱ√ITI “precious”. This second root ᴹ√ID had various derivatives having to do with both thoughts and desires, such as ᴹQ. indo “heart, mood”, N. ind “inner thought, meaning, heart”, ᴹQ. írima “lovely, desirable”, and N. idhren “pondering, wise, thoughtful”. Among its derivatives Tolkien also gave N. Idhril, untranslated but probably meaning “✱Desirable One”.
The problem with this last derivation is that in the contemporaneous narratives, Tolkien generally gave her name as Idril (SM/36, LR/141), which was the form of her name dating all the way in The Lost Tales of the 1910s (LT2/164). When writing the appendices to The Lord of the Rings, Tolkien again used Idril (LotR/1034).
This form of the name could not be derived from √ID, and Tolkien wrestled with this question in his Notes on Names (NN) written in 1957 (PE17/112). Acknowledging the problem, Tolkien first considered introducing a new root √IT “(great) enhancement” to serve as the basic of Idril, but he reversed himself, giving √ID “desire, long for” as a better basis for the name, saying: “Decision: [S.] íđril, Q Írilde”. But then he crossed all this through, and considered reintroducing √IT with a modified meaning “repeat, multiply” (PE17/112).
Based on The Shibboleth of Fëanor written 1968, it seems Tolkien stuck with the notion of basing Idril on a √IT, though he revised the meaning of the root again to “glitter” or “sparkle” (PM/363). As for √ID, the meaning “desire” seems to have been transferred back to √IR, which Tolkien gave with the gloss “desire” in notes on roots having to do with “beautiful” written in 1959-60 (PE17/155). Also, around 1957 Tolkien introduced a new root √IN-I-D to serve as the basis for Q. indo “mind” (PE17/155). The various words having to do with “thoughts” derived from 1930s ᴹ√ID may have been transferred to √IN-ID, a possibility suggested by Elaran in a Discord chat on January 26, 2018. If so, Tolkien may have abandon √ID, using only √IR and √INID going forward.
Assuming the above reasoning is correct, the conceptual development would be 1910s ✱ᴱ√IRI “wish, intend” >> 1930s ᴹ√ID “desire, thoughts” >> 1957 √ID “desire” >> 1959 √IR “desire”, with thought words transferred to √IN-ID (later √I-NID) and the name Idril transferred to √IT “glitter”.