Yes, I realise my presentation was too skeletal. It just takes so much time to write detailed arguments. Anyway, here is some of the evidence from the Pali suttas:
(1) At MN 44 samādhi is defined as cittassa ekaggatā, one-pointedness of mind.
(2) At MN 117 ariya sammāsamādhi is defined cittassa ekaggatā equipped with the other seven factors of the eightfold path. This definition is repeated at DN 18, SN 45.28, and AN 7.45.
(3) At SN 48.9 the samādhindriya is defined as follows: “having made a foundation through letting go, you gain samādhi, you gain cittassa ekaggatā.” Here samādhi and cittassa ekaggatā are evidently equated. This is repeated at SN 48.10, SN 48.11, SN 48.50, and AN 1.345.
(4) In the Iddhipāda-saṃyutta, at SN 49.13, samādhi and cittassa ekaggatā are again equated.
(5) There is also the expression samāhitaṃ cittaṃ ekaggaṃ (“the mind was concentrated, one-pointed”) which occurs in a number of suttas, for instance, MN 4, MN 19, MN 28, MN 29, MN 30, SN 35.134, AN 3.40, AN 3.130, AN 4.12, and AN 8.11. In all these cases samādhi and ekaggatā are used synonymously.
(6) On top of this we have the instances mentioned by Bucknell, but which he regards as late. These are MN 43 and MN 111, which mention the five jhāna factors, and SN 40.1.
(In all of the above, first jhāna is included whenever samādhi is mentioned, since it is part of the definition of sammāsamādhi.)
It is possible that Bucknell’s unease with ekaggatā as a factor of first jhāna is a result of reading too much into the term. At one point (footnote 10) he says that it is equivalent to cetaso ekodibhāva, the “unification of mind” achieved in the second jhāna. But this is far from clear. In fact, in the suttas the term ekaggatā is used quite broadly, even in contexts outside of meditation practice. A particularly instructive example is found in AN 5.151, where it is said one should listen to the Dhamma with one-pointed mind, ekaggacitta. This presumably means that one is continuously attentive. The same sutta further qualifies the listening as avikkhittacitta, “with a non-distracted mind”, which makes it fairly certain that ekaggacitta here means “continuous attention”. This, then, is how I would understand ekaggatā: one-pointedness in the sense of not being distracted.
The term cetaso ekodibhāva, however, which describes the second jhāna, is used much more narrowly, and refers specifically to unification of mind. (Ekodi means unified.) This is quite different from being non-distracted, and it implies the non-changing perception of a unified object of attention. It is in the second jhāna that samādhi reaches its pinnacle and that is why the pīti-sukha of this attainment is called “born of samādhi”, samādhi-ja.
This is actually more evidence than I had realised was there! Once you start looking, it is surprising what turns up.
This refers to the alternative division of the jhānas into five rather than the standard four. In the fivefold scheme the first two jhānas are instead divided into three: (1) jhāna with vitakka and vicāra, (2) jhāna without vitakka but with vicāra, (3) jhāna with neither vitakka nor vicāra. The fivefold scheme thus adds a jhāna between the first and second jhāna of the fourfold scheme. The third and fourth jhānas are the same for both schemes.
It is true that the fivefold scheme become more prominent in later literature, especially the Abhidhamma, but it is by no means missing from the suttas. For instance, it occurs in the following suttas: DN 33, DN 34 (both of which are considered late by Bucknell), MN 128 (and in its Madhyama-āgama counterpart, MĀ 72), SN 43.3, SN 43.12, and AN 8.63. I am happy to admit that the evidence here is not particularly strong, especially so since both SN 43.3 and SN 43.12 are part of a repetition series and insubstantial in content. Still, I think the evidence is strong enough to regard this as part of early Buddhism, at least until a detailed study shows otherwise.