Change hashing function used for schema subject / contents telemetry

Apparently we need to line up using MD5, not SHA.

Er, UUID-spelling of the md5 of the schema document, its references, etc. then rotored through java bytebuffer, then popped off as low and high longs fed into UUID. Read [this code](https://github.com/confluentinc/schema-registry/blob/b2fa7adb980f77d276b499ea42d462b1e09aad86/core/src/main/java/io/confluent/kafka/schemaregistry/storage/KafkaSchemaRegistry.java#L2717) and follow references from there