ensure chill URI escaping is done in all the places

### Description
Related in spirit to #6761, except with cloud object paths, based on a couple of discussion threads in #8903:
https://github.com/apache/incubator-druid/pull/8903#discussion_r348869424
https://github.com/apache/incubator-druid/pull/8903#discussion_r349756060

I believe I have fixed up all of the S3 extension, but there are a couple of potential candidates remaining not addressed in that PR that might have this issue:
https://github.com/apache/incubator-druid/blob/master/indexing-hadoop/src/main/java/org/apache/druid/indexer/JobHelper.java#L771
https://github.com/apache/incubator-druid/blob/master/extensions-contrib/azure-extensions/src/main/java/org/apache/druid/storage/azure/AzureDataSegmentPusher.java#L130

Start with the S3 URI-escaping changes discussed in PR #8903, then inspect indexing-hadoop/src/main/java/org/apache/druid/indexer/JobHelper.java at line 771 and extensions-contrib/azure-extensions/src/main/java/org/apache/druid/storage/azure/AzureDataSegmentPusher.java at line 130. Confirm whether these and any other cloud object path handling still omit escaping, and consider the work complete when all identified paths are handled consistently.

Written by the indexing model from the issue text.