That is kind of an open ended question. I think the best approach would be to look at what the requirements are for testing, and then what tool meets you needs. I cannot give a comparison (I work in SP and Rational right now) but that is where I would start. If you have both tools nothing says you cannot use them both (if the objective is to run a load test the machine you are testing does not know where the load is coming from). I will say I like the interface of SP over many of the tools I have seen or tried before.
But in the end doesn't it all come down to BEER? Beer is the ultimate answer to all questions in the universe so yes the answer to your question is BEER.
To start with the financial aspect, both Silk Performer and WebLOAD are licensed tools but the license cost of WebLOAD is far less when compared with WebLOAD. But with regards to features, Silk Performer would far exceed WebLOAD in the foll aspects -
1. The variety of protocols it supports is vast compared to WebLOAD which as far as I know supports HTTP, HTTPS, FTP and a few other protocols.
2. As Larry says, Silk Performer's interface and scripting capabilities are much better.
3. The analysis capabilities provided by SP exceed that of WebLOAD.
There are a lot more advantages, but of course if you can afford the best tool, go for Silk Performer, but if you cant then do a Proof of Concept with WebLOAD and if it fits the bill, go for it.
i personally would suggest go for webload as i gives almost all the features of the silkperformer or LoadRunner,if it is web application(i mean plain web application).Now coming back to analysis part,if you are good at interpreting the results, then you can work with any tool no matter in what format you have logs.
Please note that i am not saying that LoadRunner or Silkperformer are not suitable,but yes they are pretty costly compared to Webload.