1.Cils are independent graphically.
[It is true for 6 Cils. But no glyph for Cil-Y is discovered yet, though many authentic texts mention about it.. and there are more consonants like T, D, ZH which have ‘Cil’-ing behaviour also]
and if this ‘graphical’ independence is to be considered, it applies to the reph, prebase/postbase/belowbase signs also.
2.Atomic encoding of Cils can solve the polyvalent nature of RA/RRA etc.
[But it cannot solve the similar polyvalent nature of its pre-base RA-sign]
3.It is INCORRECT to write njaayaRRu with njaayaRu (cil-R derived from RA).ie the fallback of njaayaR will be njaayaRu if no Cil-R is present in the font.
[Is it correct to write paampu (പാമ്പു്) with the spelling പാമ്പു് ( പ ാ മ ് പ ു ് )?, the readability in this case is worser, if the font doesnt have a mpa(മ്പ)]
4.The stripping of ZWJ creates invalid constructs like PonveeNa (പൊന്വീണ). Cil-encoding solves this problem.
[the stripping of ZWNJ also creates invalid constructs (കൊയ്രാള, മൊയ്ലി,സമ്പദ്രംഗം). It is the Cil-ing behaviour of YA and DA here makes the problem, but as no ‘independent’ glyph is available, atomic encoding of these Cils is not possible and hence not solvable)
Moreover the Cil representations that are intended to be derived are already derivable using the current joiner methods. If there are problems with the joiners, measures to solve those problems must be taken, not “workarounds” like these. As the joiners and their ‘problems’ are common to many languages , a common solution is to be adopted to solve it.