The Tx that Merchant A receives may or may not be heard about by Merchant B. It is unwise IMO to make a security feature rely on the assumption that all nodes will know about all unconfirmed Tx.
Merchant A of course rebroadcasts the tx himself to make double spends less likely to succeed. As long as no block was found this is a similar situation to what we have now. Once the first tx is in a block, well, the merchant really should make sure he gets that. I have no info on this but I would expect network permeation to be very high.
It is not a security feature but a "safe enough" feature. This means it does not have to be perfect but just 98% or so, way better than what we have now.
Then, then the buyer mines his own block sending the coins back to himself. Merchant B will accept zero-conf Tx building off that block if he doesn't know about the Tx from Merchant A.
? The sending coins back tx conflicts with the tx to Merchant B so he will never accept it if he has seen the block. Also note that as soon as the block with the malicious "sending coins back" becomes public it is certain a miner will claim the payment - because the first tx is public, too.
The buyer could peer directly with Merchant B so Merchant B is likely to see the new block faster than Merchant A.
Are you talking about a Finney attack? The risk of orphaning the new block would be much too high.
This gives the buyer more time for his zero-conf Tx to be accepted by Merchant B. Merchant A must "be on the ball" to quickly realize the dbl-spend and make sure all other nodes know this.
see above
Naturally or purposefully occurring 1-2 block blockchain forks may make this proposal unreliable as well.
These are so rare they do not matter and do not hurt this proposal in any way.
Also, what if there are buyers who always dbl-spend coins, so that merchants will be discouraged from accepting zero-conf Tx using this method? There is no additional cost to buyers who dbl-spend in this scheme anyway.
This is exactly what will happen with the current system. "double spending destruction" removes the incentive to do it so I doubt a lot of people would go through the (maybe very small) trouble. Also most people still have a conscience without financial gains. Cunicula's suggestion even addresses this with a penalty.