The practical differences between tenses
The McGraw-Hill Handbook of English Grammar and usage, p.81, says:
"The simple tenses (generally) refer to actions that take place at a single moment in time…The three perfect tenses, on the other hand, deal with actions that span a period of time."
This is complete rubbish, isn't it?
If I say "I walked to the station" or "I have walked to the station" there is absolutely no difference in the time span of the action. The difference is really in the narrative viewpoint.
"I boiled an egg"--"I had boiled an egg"…absolutely no difference in the time span.
And when we say (as the book itself details) "I walk to the station", we are generally referring to recurring or habitual actions, which is the complete reverse of "actions that take place at a single moment in time".
Yes, it's mostly rubbish.
To start with, the phrase "the three perfect tenses" is already pretty meaningless. I suppose the author is thinking of had done (the "past perfect"), have done (the "present perfect") and will have done (the "future perfect"); but there's nothing really special about those three as opposed to having done, to have done, may have done, can't have done, should have done, and so on.
Secondly, even if we set that aside, it makes little sense to talk about the semantics of had/have/will have done as opposed to those of did/do/will do, because had done doesn't actually have the same sort of semantics as has done. For example, whereas "She says he's lived to the age of 95" doesn't work if "he" is someone who's died, there's no problem with "She said he'd lived to the age of 95" even if "he" is someone who had died long before. This is because had done serves both a role along the lines of do : has done :: did : had done and a role along the lines of do : did :: did : had done, so "She said he'd lived to the age of 95" can simply be analogous to "She says he lived to the age of 95".
Thirdly, even if we give the author the benefit of the doubt and assume that (s)he was generalizing unthinkingly from did vs. have done, the key difference is that have done requires a relationship to the present, whereas did presupposes a specific past time of interest. This relationship to the present can be because the action continues to the present, which of course would imply that the action spans a period of time; but firstly, the relationship to the present is just as often that the action produced a state that continues to the present (e.g. "I've already eaten" meaning roughly "I'm not up for eating a meal right now"), or that the action is repeated multiple discrete times (e.g. "I've been there twice in the past three years"), and secondly, any event that can be described using have done will instead be described using did as soon as it no longer has this relationship to the present. So as you say, it's just a difference in the narrative viewpoint.
Many people would say Paul Roberts is ancient now, but there are things that do not change about people, and "Understanding Grammar" has some good observations.
"We shall use the term notional time to mean time as we think of it, perhaps naively, in the physical world. We shall use the term linguistic time to mean the linguistic expression of notional time."
I take the phrase "perhaps naively" for an abreaction to classicist approaches, here is full text.
https://archive.org/stream/understandinggra013304mbp/understandinggra013304mbp_djvu.txt
Reading many grammar books, I came to think the same, on what grammar books cannot do: they cannot give rules for thinking. We have to do our thinking on our own, always. However, this is not a complaint. :)
It is absolutely fine that there is not and there cannot be a grammar rule to decide if we want to say we live somewhere, we are living somewhere, we have lived, or we have been living somewhere. Our phrasing will depend on how we perceive our staying in a place.
We have four Aspects, to tell that.
Human grammar is not separate from human living and thinking. We can associate our grammar and natural human mapping, as with geography and travel. We people live on Earth. We usually view land or seas as extents. We give at least psychological borders to areas in which we are. We perceive routes and ways to places. We happen to be at landmarks and places.
We can use these natural mapping variables for grammar.
ON: I live here (plain statement ON a cognitive map)
IN: I am living here (I focus on things IN a time span)
TO: I have lived here (I think as about things TO a time)
AT: I have been living here (I combine my picture of living TO a time with that for withIN a time).
As for people and things changing, Anaximander proposed conceptual abstraction already in ancient Greece.
https://en.wikipedia.org/wiki/Anaximander
Humans have not changed much, with regards. :)
Feel welcome to tell me what you think about my idea,
https://travelingrammar.com/2016/04/19/language-mapping-summary/