In many course books for learning English you will find the 'rules' for reported speech (in Esperanto course books too).

But as Lewis points out in 'The English Verb' these reported speech rules are not the whole story and any combination of tenses may lead to natural sentences in English.

He says he is going to buy one
He says he was going to buy one
He said he is going to buy one
He said he was going to buy one.

Kiel oni esprimus en Esperanto tiujn kvar frazojn?

sudanglo:He says he is going to buy one
He says he was going to buy one
He said he is going to buy one
He said he was going to buy one.

These are subtly different things, I think.

He maintains that he will buy one.
He maintains that he had at one point planned to buy one, but whether he did or not is irrelevant.
He said he will buy one, but he might have changed his mind by now.
He said he had planned to buy one, but whether he still plans to, or if he already bought one, is irrelevant.

I'd translate these as:
Li diras, ke li aĉetos unu.
Li diras, ke li planis aĉeti unu.
Li diris, ke li aĉetos unu.
Li diris, ke li planis aĉeti unu.

I'm certainly no expert, but I thought I'd participate.

sudanglo:
He says he is going to buy one
He says he was going to buy one
He said he is going to buy one
He said he was going to buy one.

Kiel oni esprimus en Esperanto tiujn kvar frazojn?

Li diras, ke li aĉetos/estas aĉetonta unu.
Li diras, ke li estis aĉetonta unu
Li diris, ke li aĉetos/estas aĉetonta unu
Li diris, ke li estis aĉetonta unu.

Close?

"Li diris, ke li planis aĉeti unu."

Would this be also a possible translation for: "He said he would buy one"?

or "Li diris, ke li estus aĉetonta unu." would be more apropriate?

There are more issues here than just the implication of choice of tense in reported speech.

Both French the previous international language and English the current international language have the means to express the idea of going to - which you might summarise as okazivo antaŭvidata. Whether Latin the earlier international language could also express this idea I don't know; perhaps some classical scholar could let us know.

There is also the issue of whether unu is the appropriate choice for one here, with its emphasis on quantity.

A wide range of causes can be responsible for the speakers sense of antaŭvido.

She's not going to like that (I know her)
It's going to rain (dark clouds approaching)
I'm going to buy one (intention/decision made)

Estas/is X-onta is quite commonly used in Esperanto judging by the some 350 hits in the Tekstaro, but it is not clear that it equates exactly to going to. Often the meaning appears to be on the point of, or about to, or referring to some unfolding event (= was to or is to) . Sometimes the natural English translation would be going to but not always.

Maybe if somebody announces (or you report it), that they are going to buy something, the simple future aĉetos is adequate. You could colour this up a bit with intenci or decidi. Mi estas aĉetonta seems strange, but mi estis aĉetonta plausible in the appropriate circumstances.

As a sidebar: do you find it strange how native speakers of English unerringly choose a particular verb form, but 99 out of 100 could not explain, for example, when to use have X-ed rather than X-ed, or shall rather than will.

Sometimes there don't even seem to be the words to describe economically a particular verb usage. How can it be that a language makes certain distinctions but doesn't have the vocabulary to identify those distinctions?

sudanglo:As a sidebar: do you find it strange how native speakers of English unerringly choose a particular verb form, but 99 out of 100 could not explain, for example, when to use have X-ed rather than X-ed, or shall rather than will.

Sometimes there don't even seem to be the words to describe economically a particular verb usage. How can it be that a language makes certain distinctions but doesn't have the vocabulary to identify those distinctions?

I think most people would be able to tell the difference between 'have X-ed' (the perfect) and 'X-ed' (the simple past), but you're right that it's not always easy to describe why you'd use one over another, it's just a case of convention (e.g. French uses its perfect all the time now, it's completely jettisoned its simple past except in formal written works). The distinction between 'shall' and 'will' is different because it's not really followed any more (if it ever really was, except in prescriptivist grammar books), though traditionally 'shall' was restricted to the first-person except for emphasis.

it's just a case of convention

Decidedly not. There's a meaning difference, an expression of different perspectives on the past (it's to do with finished and unfinished time).

The strange thing is that even when you have seen the difference articulated and seen how it fits a whole range of sentences using the present perfect or the past, there is still a sense of wonder of how you can be regularly and most consistently expressing this difference without any explicit awareness of what you are doing.

It is no surprise that you cannot tell someone how to cycle or to walk. These muscle skills are definitely in the domain of the unconscious, but with language you feel it should be different.

It invites the question as to whether there are features in Esperanto which experienced speakers employ, but are unable to correctly articulate or explain.

It is noteworthy, in respect of many aspects of the English verb that are in fact used systematically and consistently by native speakers, that false theories of the usage abound. So that certain cases are considered exceptions to the (false) rule, or that there are several different uses, when in fact all can be encompassed by a single explanation.

The distinction between 'shall' and 'will' is different because it's not really followed any more

Oh really! Then no difference between What time will we arrive and What time shall we arrive!

sudanglo:The strange thing is that even when you have seen the difference articulated and seen how it fits a whole range of sentences using the present perfect or the past, there is still a sense of wonder of how you can be regularly and most consistently expressing this difference without any explicit awareness of what you are doing.

It is no surprise that you cannot tell someone how to cycle or to walk. These muscle skills are definitely in the domain of the unconscious, but with language you feel it should be different.

It invites the question as to whether there are features in Esperanto which experienced speakers employ, but are unable to correctly articulate or explain.

It is noteworthy, in respect of many aspects of the English verb that are in fact used systematically and consistently by native speakers, that false theories of the usage abound. So that certain cases are considered exceptions to the (false) rule, or that there are several different uses, when in fact all can be encompassed by a single explanation.

I suppose that is why Noam Chomsky has been unwilling to consider Esperanto a real language. He means it has not yet had the time to settle down enough to let those unconscious rules, attended by experienced speakers, dominate over the rules that Zamanhof was aware of and managed to put in writing. It seems as if Chomsky only considers the unconscious rules really worthy of the linguists' attention.

Rugxdoma:

sudanglo:The strange thing is that even when you have seen the difference articulated and seen how it fits a whole range of sentences using the present perfect or the past, there is still a sense of wonder of how you can be regularly and most consistently expressing this difference without any explicit awareness of what you are doing.

It is no surprise that you cannot tell someone how to cycle or to walk. These muscle skills are definitely in the domain of the unconscious, but with language you feel it should be different.

It invites the question as to whether there are features in Esperanto which experienced speakers employ, but are unable to correctly articulate or explain.

It is noteworthy, in respect of many aspects of the English verb that are in fact used systematically and consistently by native speakers, that false theories of the usage abound. So that certain cases are considered exceptions to the (false) rule, or that there are several different uses, when in fact all can be encompassed by a single explanation.

I suppose that is why Noam Chomsky has been unwilling to consider Esperanto a real language. He means it has not yet had the time to settle down enough to let those unconscious rules, attended by experienced speakers, dominate over the rules that Zamanhof was aware of and managed to put in writing. It seems as if Chomsky only considers the unconscious rules really worthy of the linguists' attention.

Aside from the fact that Noam Chomsky is an ignoramus who speaks on the subject about which he does not have basic knowledge, his theories have not been proven so far.