Very different.
It's a centralized multi-sig service that works with 'oracles': "Oraclize takes care of providing the 2nd signature to your partially signed transaction."
There are many project like this, and I actually think the term "smart contracts" does not necessarily apply here. The contracts still have a mediator (Oraclize), and are not programmable/customizable. I would call it a "multi-signature oracle service" or something like that... The potential uses for this are numerous, but greatly exaggerated in variety.
The point of Counterparty 'smart contracts' is to be able to create entirely custom decentralized applications from scratch, on a fully peer-to-peer system. I'm actually thinking we could refer to this as dApps, programmable money, or something similar until people learn the difference. I can't even count the amount of times people were underwhelmed by the idea of Counterparty smart contracts, because they've seen dozens of projects with basic multi-sig acting as middlemen and have become convinced that that is a 'smart contract'. It's not... at least not really.
Counterparty/Ethereum mercilessly blow these out of the water since you are able to code new features.
Once you see the difference here, it's really hard to unsee:
- 'Buzzword' smart contracts are multi-signature/Bitcoin services which are mediated by some 3rd party/website. They are not fully decentralized, and cannot be programmed.
- 'Real' smart contracts support turing-complete scripting for making your own, fully or partially decentralized financial instruments and tools.
In fact, the first can be done using the latter. But absolutely not vice versa.
Note that I'm not saying that Oraclize and others are not providing a great service. They very well might be, and in many cases this is extremely useful. However, I think we need some sort of way to distinguish between distributed, programmable contracts and static, semi-centralized ones.
I hope this makes sense...