Encryption of the communication would prevent MITM snooping but it is not at all easy. The problem is the initial key exchange. The two nodes have to exchange a key first to use for the encryption/decryption. This key exchange itself can't be encrypted and the ISP (aka the man in the middle) can swap the key with its own.
The reason why https works for websites is that your computer is relying on a centralized "master key" (CA) to validate that the key that the website's server sent you is not changed by the ISP or MITM. We can't add that to a decentralized network such as bitcoin.
BIP151 has an encrypted traffic exchange in place but it doesn't specify a method for the initial key exchange as well so it'd have to work on the premise that the first connection to the other node is trusted. I think BIP150 covers that part but I've seen a few counterpoints against that proposal.
Traffic analysis should be slightly harder against encrypted streams but it should still be fairly easy to identify them; periodic spikes in traffic, etc. If the censorship does happen, running it within Tor would be sufficient since the user would probably have to use Tor or something else to bypass it in the first place.