I don't know Bill, not very technical about this, what's your perception?
My guess is the blockexplorer is using the "patch" or if its not, its listing the blocks. What needs to happen is the consensus model should decide which block is correct.
Before under the old website there were duplicates, but not as bad. I assume the database is copying the information, and not deleting it later when the wallets reorganize. Normally this is rectified when the wallets walk back to the duplicate decide which one is correct, and then begin rebuilding the chain. Any transactions other than mined blocks are added to the blockchain so no transactions are lost. (except for mined blocks).
Breakcrypto said his patch does not leave duplicates in the blockchain, but it does prevent the rewind attacks as he calls it.
I want to beleive that his solution works, but I'm still in need of evidence. My experience with testing the patch was there was no reorganize.
This is crucial because in order to have only one correct blockchian wallets need to have consensus and therefore duplicate blocks are eliminated. Only one blockheight is allowed.
If his patch eliminated the orphans when mining then the only logical explanation is that his wallet is operating on a fork. an easy way to verify this is to look at the txid of the generated block and look on block explorer. If there are multiple blocks at that height, then its on a fork.
The reorganize is the correct response.