... when deciding what Performance rating to give on a review.

A good review system will have some abstraction from the final score. The tendency for a 1-10 score to be polarized higher than it should be means that you're really not getting a good picture of actual performance. There's too much bias in just choosing a final, concrete score.

Here's the solution:  
In short, Itemized service reviews and aggregated total scores.    
   
For each of the services rendered, the provider gets a 1-10 score.  
This means that, if the provider performs DFK, BBBJ, CFS, they would get 3, 1-10 scores for their work.  
Then, the scores are averaged to determine the aggregate performance score.    
   
   -Performing more services would allow a provider to pad their scores with additional services only if the services provided were actually worthy of high scores.  
   -Performing a limited number of services would allow a provider to score higher only if said services were performed well, but would penalize a provider more if any of those services were performed poorly. It does not explicitly penalize providers for not performing certain services.  
   -Additional data capture would allow the search function to be more granular. Clients would be able to search for who was performing well at a particular service.  
   -Reviews would be more useful to providers because they would have granular feedback on which services they offer score highly and which services they could work on.