Most of the information you'll find in the performance trace actually is internal information where you won't find much documentation about.
This is partly due to the fact that the way SAP HANA processes queries on this level had been substantially changed (and will be changed further) over the course of relatively few releases.
By the time any documentation would be ready and complete, it would be outdated.
Another aspect is that besides things like "TempIndex" which point to temporary result sets that get materialised during the query execution, it is typically rather not obvious what a specific processing step does and whether it's execution time has a critical impact to the overall query execution time. The naming of the Operators is not straight forward if you're not the developer of them.
About your other questions:
"overallQoExecution" - QO typically stands for query optimizer (column store level), so I'd say this is the total time spend with this piece of code
"Inverted index: not used but present" - is just an information and not actual data processing. It means, the column has an inverted index, but the optimiser did not find it beneficial for this query, so it isn't used this time around.
"deleteTempIndices" - clean up of intermediate result sets (don't think secondary b*tree index, but think temp table here!). Nothing that you could influence, except for changing your query to eliminate intermediate result sets in the first place.
Ok, not sure if this helps you all that much further.
As a general hint, I'd say it's better to start off with the bigger picture of the query execution and not to focus on single operations. That's because these single operations always occur in a set of operations and that is what mostly drives how much data needs to be processed.
So looking for where and how much data is used and forwarded to the next processing step is an approach to look at the query processing.