I don’t see exactly what you are disagreeing with, as you seem to accept what I’ve said - that ‘chandas’ in the passage refers specifically to versified early-Vedic, and not to any other kind of Sanskrit (in general).
The word chandas (when used to denote a language in all the BCE texts I’ve read) invariably means metrical early-Vedic, and no other so-far unattested archaic gathā-language has come to light in any BCE textual source, except gāthic-Avestan or Old-Persian (one or both of which however I accept Pāli as being influenced by) - but Avestan and Old-Persian were never called/equated-to chandas in mid-1st-millenium BCE India to my knowledge.
Further, in hundreds of sūtras of Pāṇini (circa 4th century BCE) in which he explains the grammatical peculiarities of the ‘Chandas’ language (i.e. as opposed to bhāṣā or ‘spoken classical sanskrit’ of his time), the word Chandas means no other language, other than early-Vedic. Wherever he says ‘chandasi’ (either explicitly or by anuvṛtti), those rules fit the description of early-Vedic. This is what Kātyāyana, Patañjali, and practically everyone else in BCE (whose texts are extant) take the word to mean. So I don’t see what else Geiger could have meant by “archaic Gātha language” and how an unattested phantom language would plausibly connect to the term ‘Chandas’.
The Buddha’s resistance to the “chandaso āropema” suggestion is not however that it would prevent “conversion” (which is also an idea foreign to the first-few generations of Buddhists, as most of the society in the time of the Buddha, lay or monastic, didn’t consider themselves systemically ‘converted’ by the Buddha or by his non-Buddhist contemporaries, but rather persuaded by the Buddha’s eloquence, wisdom and worldview. There isn’t anything in the EBTs that I see to indicate that they considered the Buddha’s wisdom to be uniquely Buddhist, to which they had to convert or bear exclusive alleigance/loyalty to, and the term ‘conversion’ would imply such exclusive-alleigance, and an explicit rejection of what they formerly had alleigance-to).
The explicit reason for the Buddha’s resistance (that is mentioned in the ‘sakāya niruttiyā’ passage) is that such an effort wouldn’t “add to the clarity of those Buddhist monks whose knowledge of the Buddhavacana was alpa” (“n’etam… appasannānaṁ vā pasādāya”). He thought that if the Buddhavacana wasn’t understood already (in its pre-existing form), it would not be made any clearer (for those who couldnt grasp the Buddhavacana properly) by elevating it to early-Vedic verse.
Therefore there is no Sanskrit vs Prakrit distinction (or a Sanskrit vs. Pali distinction) meant in the passage - moreover when the Buddha heard a sa-svara recitation of the aṭṭhakavagga by Soṇa Kuṭikaṇṇa he appreciated it - and a sa-svara recitation of a text that is not normally recited with svaras (pitch accent) was only possible in late-Vedic as I’ve argued in that post (linked above). So the distinction here is between late-Vedic (i.e. classical sanskrit) and versified early-Vedic. This incident is mentioned both in the Pali canon and the Sanskrit tradition underlying the Divyāvadāna. The buddha there is reported to have praised the sa-svara recitation saying it was well understood, well remembered and well-recited by Soṇa: “sādhu sādhu, bhikkhu, suggahitāni te, bhikkhu, soḷasa aṭṭhakavaggikāni sumanasikatāni sūpadhāritāni". The epithet ‘koṭikarṇaḥ’ (sanskrit for kuṭikaṇṇa) used for Soṇa literally means “one with 10 million ears” (perhaps referring to his superior listening abilities and exact recitation of what he heard).