According to Schneier doing this is a bad idea - (or so I recall from the A.P. book which I've not reread in quite a while - I may be wrong) if you use the same (or similar) cypher. i.e.:
blowfish(blowfish(plaintext,key1),key2) is bad, but rsa(blowfish(plaintext,key1),privatekey) is ok.
Does it apply even if it is the same cipher but with different key length and/or block size? I was pondering such "encapsulation" for the situations when The Government forbids using ciphers stronger than <limit>. Then use as strong one as you wish, and encrypt the result in the legally-weak wrapper. Once they ask for your escrowed keys, or bruteforce it, they will figure out that you are a crypto-lawbreaker - but you will pass a routine automated screening. And once you catch their interest, you already have problems anyway.