looks like so much hashrate has dropped from corechain, and jumped onto blake2b, that blake2b will be the longest chain very shortly
Not really. All old clients halted at block 961,639. Only upgraded ones follow the Blake chain, which is a hard-fork.
If you have different network rules, then it is easy to have the longer chain, when you count only block numbers. But even if it will be longer, then it won't reorg BTC, because of being incompatible. And it probably won't be longer than eCash, which also didn't reorg BTC, and just dropped the difficulty into one:
https://explorer.alpha.ecash.ninja/thousands of people dumped that legacy chain
And for some reason, it is still hard to find anyone, willing to sell legacy coins. I wonder why, my BIP-110 coins are waiting for buyers, but it is hard to find any.
why is the legacy chain slowing down and the new chain doing so well?
Because if you do a hard-fork, then you can change anything, and produce millions of blocks, without any effort. Also, your post is a proof, that you blindly trust the block explorer, because if you would have a BIP-110 full node, then you would instead complain about it being halted at block 961,639.
thousands of people signaled for bip110
If counting node numbers would be sufficient for consensus changes, then we wouldn't need Proof of Work. It would be "Proof of Node Count" instead, and new rules could be decided by just counting different IP addresses. But this is not how it works, because it is much easier, to allocate many IPs, than to produce many blocks.
why are blocks taking so long on the corechain, then?
They are produced as usual, every 10 minutes.