Gordian knot conundrum

The basic question that I try to find an answer to is this:
Is it pure chance that methionine is the universal start amino acid in protein synthesis or is there a logical explanation why this amino acid and not any other (formyl-methionine doesn't count and glutamine or isoleucine are much less efficient at initiation of translation than methionine) is used.

Francis Crick wrote:

"Discussion of the actual amino acids used in the code may not be very profitable. Some less common amino acids, such as cysteine and histidine, would clearly seem to have an advantage because of their chemical reactivity; but whether, say, methionine could be justified in this way seems less obvious."
Crick FH. The origin of the genetic code. J Mol Biol. 1968 Dec;38(3):367-79.

That may very well be true, but I think just opting for the "accident" solution is not a very satisfying answer and to have a logical explanation is better than not having any at all. Of course the structure of the initiator tRNA is the deciding factor in this, but it might have been conceivable that other initiator tRNAs might have been successful. If methionine is nearly always the first amino acid in a protein sequence, but often cleaved off with other N-terminal amino acids, then it might not be essential for the structure/function relation of the protein but still play an important role outside protein synthesis, which is where S-adenosylmethionine (AdoMet) comes in. S-adenosylmethionine is used in transmethylation (of DNA, rRNA, tRNA, mRNA-caps, proteins and lipids), transsulfuration, and aminopropylation (polyamine synthesis of spermidine and spermine). Could the fundamental processes of translation, transmethylation and polyamine synthesis be tethered or is it merely an interesting coincidence, a correlation in the absence of a "cause-and-effect" relation? How could one adress this question? How do you untangle a Gordian knot without using brute force that would wreck the whole system?