The recommendations will vary between different types of VDP deployments.
For example, if VDP is deployed in a server with a small set of cores, this may not allow sufficient capacity for thread concurrency. The reason this is an important factor is because if you decide to run scheduler jobs with a high concurrency level, these queries will be executed against the same VDP server even in a cluster environment, so you can run into extraction issues if you do not have enough cores and processing power.
Another factor is the type of queries you are running. You should not have many issues running operaional queries against the server whereas analytical queries will have a much higher impact on the server.
I recommend you read through this KB Article for more general recommendations regarding the processor: