In the grand sum of things it just a red herring.

There is an attack possible via miners itself to attack segwit address which itself very theoretically.
It would require a cartel of 51% of miners. This is basically a 51% attack similar to Reorg. I might not like bitcoin(dot)com but they do have the view point you are talking about:
https://news.bitcoin.com/risks-segregated-witness-opening-door-mining-cartels-undermine-bitcoin-network/
So incense it basically a 51% attack on segwit should be the least of your worries.
CSW to be fair is also the source so alot of points are exaggerated. For exp, a HF is needed to make this happen. The HF required to reverse Segwit would mean they would be mining a valueless fork with no economic node support. Alot of IF need to happen to make this iffy idea real.

The reason segwit works by dividing between address is pretty safe.
It hides its increased block size by changing the definition of a block to be measured as one million "units" instead of bytes. The "witness" signature data would be separated from the Merkle tree record of who is sending or receiving the bitcoin. The "witness" data is moved to the end, and each byte of it would only count as one quarter of a "unit". The overall effect would be changing the average block size to about 1.8 MB instead of 1. This means the existing bitcoin protocol doesn't change, allowing it to work without as much upgrading of software.

It also addresses signature malleability, by moving signatures out of the transaction data, making them impossible to change.
Old nodes validate them because from their perspective these transactions don't require a signature at all (and they don't see one), and new nodes validate them because the required signature is in the Segregated Witness. And since both old and new nodes hash the transaction data into the same transaction ID, everyone agrees on the makeup of blocks, and, as such, on the structure of the entire blockchain
https://bitcoinmagazine.com/articles/segregated-witness-part-how-a-clever-hack-could-significantly-increase-bitcoin-s-potential-1450553618/
https://bitcoincore.org/en/2016/10/28/segwit-costs/