@Merry Andrew,
I'd say the etymology for 'inflammable' points to it having a close connection to 'inflame'. The 'in' here is necessary for the particular meaning of 'inflame but it clearly isn't necessary for the adjective form.
I don't think that the "redundancy" is any big deal, for language is chock full of redundancy. Even though some substances are equally cap
able of being 'inflamed' or being 'flamed', the former is a more accurate description of this particular process.
[I wonder, was there ever a 'flame something' which had the literal meaning of 'put flame to something' or has it only been used figuratively and recently in 'flame someone'?
However, given that 'flammable' works independently from 'inflame' illustrates that though there may be an etymological connection, different word forms have different meanings.
How do they acquire such meanings? Not from the mistaken notions of "logic" as put forth by folks like OmSigDavid, but by the people who use them in language and infuse them with new meaning or a slightly different nuance.