After reading the job result with maxRecords = 10000, the API sends Sforce-Locator in the response headers which can be used to fetch the next batch of records. 
To read 'n' batches in parallel I need 'n' Sforce-Locators beforehand.

• Is there a way to get these 'n' locators given the maxRecords without actually waiting to download the content?
• Is there a limit on the number of concurrent connections to this result API?  If not, Can I trigger any number of GET requests in parallel to download the result without facing connection reset issues from the server?

After reading the job result with maxRecords = 10000, the API sends Sforce-Locator in the response headers which can be used to fetch the next batch of records. 
To read 'n' batches in parallel I need 'n' Sforce-Locators beforehand.

• Is there a way to get these 'n' locators given the maxRecords without actually waiting to download the content?
• Is there a limit on the number of concurrent connections to this result API?  If not, Can I trigger any number of GET requests in parallel to download the result without facing connection reset issues from the server?