41% of my skip traces came back with a working number. Is that good?

Two different numbers are getting mashed together in your 41%. Hit rate is the share of records where the vendor returned any phone number at all. Connect rate is the share where a human answered you. Vendors quote hit rate and it's usually 60 to 80% on absentee owners. You're quoting connect rate, and 41% connect on a nine day dial window is on the healthy side of what I've seen.

Run the second vendor. It's called double dipping and it's normal, different vendors buy from different sources so the overlap is partial. I get another 10 to 18% of usable numbers on a second pass for the same kind of cost. At $96 that's cheap information even if it comes back poor, and then you'll know whether the list or the trace was the weak part.

What I'd also check is the 15 plus years of ownership filter. On absentee records that old, a real chunk of your disconnected numbers are deceased owners and the property is sitting in an estate. Those are still leads, they just need a different conversation and probate records rather than a phone.

Also, nine days of dialing on 800 records is not a lot of attempts. How many touches per record? If it's one, your connect rate is understated by a wide margin.