It depends on the reason why is your patch running slow.
You can expect a patch that simply requires lots of maths to run 3.5-4x faster on Daisy. OTOH, if patch was accessing SDRAM too much, it would be just a little faster, since that can kill performance just as much due to access latency, but on Daisy’s MCU you have cache that can help to some extend.
There’s also a special case for patches that had to access SDRAM a lot on F4 MCU, but fit on SRAM on H7. This gives a huge performance increase, I wouldn’t be surprised to see such patch run x6-x8 times faster.
If you’re really curious about performance of a particular patch on Daisy, I can run it as a test, because OWL runs on Daisy Patch too. And there’s Daisy-only gen~ integration to which you can port your patch, but you’ll have to ask someone else to test it.