OMFG awesome, yeah?!
Ok, we all know SciFinder lies sometimes. It's not always their fault, sometimes the reaction they've faithfully indexed and you've faithfully, hopefully, please-Lord-let-this-one-work-ully set up never really gave the yield the authors said it did in the first place.
But this is just taking the piss.
Look at the responses on Twitter and in the comments of the Synthetic Remarks blog linked above. Their repeated refusal to admit that the example quoted is total nonsense, indeed their apparent refusal even to cast an eye over it, speaks very poorly of their commitment to accuracy and to correcting errors, which I would expect to be at the heart of their business.
The CAS registry and SciFinder contain a huge amount of important data. Please reassure us that you value its accuracy.