What about those that use the memo for another purpose like the ecommerce plugins? Will they be affected
I'm not sure if this was directed at my proposal or not. But with the protocol I propose you could either have 18 characters to work with in the memo and the receiver could have an observer who they can pull relevant transactions from, or you could have the full 51 characters but the receiver could have no observer, meaning you would either have to send the funds to the receiver publicly or they would only be able to receive the transaction if they scanned the blockchain for TITAN transactions (this is ignoring the case where the transaction's existence could be notified to them by the sender through methods outside of the blockchain, e.g. BitShares mail).
If one is sending a TITAN transaction to an ecommerce service, we can expect that the service will be a full node scanning the blockchain for transactions meant for it. In that case, there is no need to register that account with an observer and therefore all 51 characters of the memo could be available to the sender for whatever purpose. If this is a transaction the ecommerce service is sending to the user of the service, then the service would only have 18 characters to work with in my proposed protocol (which seems like more than enough to generate a code that is unique for that service).
Of course this is all assuming we want to work within the constraints of the existing system instead of modifying it to give transactions the flexibility of specifying this observer information as additional metadata outside of the memo field.