If you consider {le mi xunre cukta}, {mi} does not function as just
another tanru component, so that it modifies {xunre} and the result
modifies {cukta}. Rather {mi} modifies {le xunre cukta}.
How about {le mi cukta xunre}, where I don't want {mi} to modify the whole {le cukta xunre} but only {cukta} (so, "the redness which is peculiar to the book of mine" rather than "the book-specific redness which has to do with me")? Is it an explicit sumti linking which I have to use in that case: {le cukta pe mi xunre} or {le xunre co cukta pe mi}? You said {le mi cukta} & {le pe mi cukta} & {le cukta pe mi} are identical in meaning where {mi} functions invariably as a kind of infix to modify the
whole sumti including {le}; then it should be impossible for the {mi} in {le cukta pe mi xunre} to target exclusively on {cukta} instead of the whole {le cukta xunre}. If it is that {mi} do actually modify {le cukta} in an exclusive way here, the remainder (xunre) is deprived of the descriptor {le}, divorcing from it, isn't it?