Are WPRs more effective at certain tracks?

There's no reason to think ratings will be more effective at some tracks than others. We could observe differences in results in a given period, but they are almost certainly coincidental from natural variance.

The other factor to consider is that the way the ratings present the chances in a race depends on lead-up runs that, in many cases, are from somewhere other than the track being raced today. That makes it difficult to form a logical argument as to why the ratings could legitimately perform better at some tracks than others (for reasons other than coincidence.) 

You could theorise that a top-rated horse coming off a lead-up at the same track as today might perform better than other top raters, but even that doesn't hold true when you look at a large sample size. Our advice would be to ignore the track of the upcoming race as a variable unless doing detailed form and considering why a particular track or expected track pattern may suit one horse more than another.