In a discussion of Noldorin versus Vanyarin phonetic developments from the late 1950s or early 1960s (PE17/129), Tolkien said in a marginal note:

In Q. both rv (< rb) and rw existed: both in EQ [Exilic Quenya] > rv. {In Q. both lb and lw > lw: both in EQ > lb [deleted and replaced by]} In Q. lb and lw existed, then both became lv. Consider[ed] correct but lv often to lb, which is sufficiently common to be allowed as a correct variant. Thus distinct from Vanyar ancient lb from lv but lw from l + w (PE17/129).

Here it seems that in Exilic Ñoldorin Quenya lw, rw merged with rv, lv (from primitive rb, lb), all ultimately becoming lv, rv, with the caveat that sometimes lv was pronounced lb, a variant pronunciation also mentioned in The Lord of the Rings Appendix E (LotR/1121). The discussion on PE17/129 seems to have been inspired by Tolkien’s decision to change the suffix -lmo >> -lvo (genitive of -lvë) in the phrase elen síla lúmenn’ omentielvo between the 1st and 2nd editions of The Lord of the Rings. Later in the same notes (PE17/130), Tolkien states that -lvë developed from the primitive pronoun ✶we “we (inclusive)”, which became the independent pronoun ve. Tolkien needed to explain how lw > lv, and this new phonetic rule provided this explanation.

This new phonetic rule contradicts the scenario originally described in the Outline of Phonology (OP2) from the early 1950s, where lw, rw and rv, lv (< rb, lb) remained distinct in the Ñoldorin dialect:

Note: since w remained [u̯] after r, l, CE lb, rb remained distinct from CE lw, rw in N[oldorin]. In V[anyarin] TQ there was a tendency to pronounce lw, rw also as lb, rb (PE19/93).

Tolkien likewise indicates that rw, lw were preserved in the table of medial combinations in OP2, with rv, lv derived only from Primitive Elvish rb, lb:

x. r/l + v, w, y: rv, rw, ry; lv, lw, ly (rv, lv in N. = V rb, lb derived only from Eldarin rb, lb) (OP2: PE19/81).

In the Outline of Phonetic Development (OP1) written in the 1940s, there are also indications in the table of medial combinations that the distinction between primitive lw, rw versus primitive lb, rb was preserved (PE19/38). In this conceptual period, the spirantalization of lb, rb to rƀ, lƀ [lβ, rβ] (and later > rv, lb) was only realized in the Lindarin dialect [pre-Vanyarin] (PE19/46), as opposed to Tolkien’s later conception that Vanyarin restopped these sounds to lb, rb but rƀ, lƀ underwent further phonetic development to lv, rv in the Ñoldorin dialect of Quenya (PE19/93 and note #118).

However, elsewhere in ink revisions to OP2 (made in the late 1950s or early 60s?) Tolkien instead said that lv, rv could become lw, rw in Noldorin Tarquesta [TQ]:

{After r, l and the diphthong ai however w remained as such >>} After s (z), n, r, l and the diphthongs ai, oi however w remained as such, and in N. TQ the tendency was here to change v > w (PE19/72 and note #22).

There was a similar change of lv, rv (< rb, lb) to lw, rw in extensive notes on the use of the Feanorian Alphabet from the 1940s:

The earliest distinction given up was that between lv, rv (< lb, rb) and lw, rw, which early coalesced in lw, rw (PE22/45).

Thus it seems Tolkien vacillated a great deal on whether or not primitive ✶rb, lb and primitive ✶rw, lw merged in Quenya, which dialects the mergers occurred in, and what the results of that merger may have been.

Neo-Quenya: Aside from the case of -lwe > -lve, there are no clear examples of this phonetic rule elsewhere in Tolkien’s writing, and in fact there remain plenty of examples of unmodified rw and lw. Where rv and lv do appear, they can be easily derived from primitive ✶rb, lb instead. For purposes of Neo-Quenya, it is probably best to assume these lw, rw / lv, rv mergers were a transient idea and ignore them.

This means we still need to justify -lve, but this can be explained by analogy with the independent pronoun ve. These ideas for Neo-Quenya usage came out of a discussion with Shihali on the Vinyë Lambengolmor Discord Server (VLDS) on 2018-01-28.