Regarding executionTime shown at explain plan

Manual presented at https://www.mongodb.com/docs/manual/reference/explain-results/ shows winning plan structure and executionStats structure , where each stage passes its resulting documents or index keys to the parent node.

I have a query about ‘executionTimeMillis’ field shown at executionStats and ‘executionTimeMillisEstimate’ shown at every stage hierarchy.

I have observed that ‘executionTimeMillis’ = ‘executionTimeMillis’ of Shard_Merge stage.
Another observation is ‘executionTimeMillis’ of shard_merge stage is maximum of all shard’s ‘executionTimeMillis’. That may be due to parallel of execution at all shards. I can understand this logic.
My query is why ‘executionTimeMillis’ of each shard is neither equal to outer stage of that shard nor sum of all stages of that shard?

Can an y one explain

‘executionTimeMillis’ is not equal to most outer stage’s ‘executionTimeMillisEstimate’ nor sum of every stage.

Can any one explain me why?

Please answer that is ‘executionTimeMillis’ of Shard_Merge equal to ‘executionTimeMillisEstmate’ of input stage?