A couple of things to consider.

First, scholarship has mostly focused on “classical” Sāṃkhya, i..e., Sāṃkhya-kārikā, etc., but this is already quite late in the development of Sāṃkhya thought, and most of the earlier literature is no longer available, aside from texts that allude to those early Sāṃkhyan models without fully detailing them, such as the Bhagavad-Gītā, Aśvaghoṣa’s Buddhacarita, etc. 

One clue we get from the Gītā’s treatment of puruṣa and prakṛti is that the radical separation between them characteristic of “classical” Sāṃkhya has not yet set in. Which implies that the version of the 25 tattvas adopted in the Yoga-system may be more in tune with earlier Sāṃkhya on that account than what became codified in the kārikās, etc. (The Mokṣadharma section of the Mahābhāratā, which is considered a late section, offers a 26 tattva model alongside the 25 tattva model, providing a universal self as the 26th, indicating, at the least, that at some point conflicting models were in play). There are a variety of differences between the Gītā’s account and the classical version, such as the classical account treats puruṣa as not only sequestered away from prakṛti is a variety of ways (it has no emotions, etc., which are all on the prakṛti side) while the Gītā’s puruṣa does feel and experience and is active; classical puruṣa is radically passive while the Gītā’s puruṣa has active capabilities, feels pain, etc.; and the classical puruṣa is radically singular (while each individual has its own puruṣa), but the Gītā’s puruṣa is threefold (kṣara puruṣa, akṣara puruṣa, paramātman; cf. BG XV 16-19); and so on. There are also items and lists mentioned in other early texts that they associate with Sāṃkhya but which were  no longer included in the classical system, indicating important aspects of early Sāṃkhya remain obscure to us.

Classical Sāṃkhya, unlike the Gītā, had a problem explaining how puruṣa and prakṛti interacted, and a theory of buddhi, the first tattva engendered by the three guṇas, reflecting the light of puruṣa so as to illuminate the rest of prakṛti. The separation between puruṣa and prakṛti became further reified when western scholars turned them into “spirit” vs “matter”, bringing in distortive baggage that utterly obscured how prakṛti was conceived in Sāṃkhya, and an example of how an eagerness to project familiar models onto alien concepts can obscure what one is observing. To quickly illustrate, which of the following would anyone in the west consider “matter” rather than “mental” or “cognitive”:

intellect, ego, mind, hearing, feeling seeing, tasting, and smelling.

i.e., buddhi, ahaṃkāra, manas, and the buddhi-indriyas.

These are tattvas #3 through 10 of the 25. (#1 is puruṣa, #2 is the three guṇas).

While the effort to attribute higher cognitive function to puruṣa by stripping away prakṛti’s cognitive functions, or the underlying cognitive capacity that animates them was already attempted in later expositions of classical Sāṃkhya in India in a variety of ways, the “spirit” vs “matter” superimposition rendered the model incoherent, and misconstrues how the tattvas relate.

The three guṇas are, at once, psychological, physical, emotional, etc., a tripartitie analogue to the Chinese yin-yang, which permeates the dynamics of everything, from food to medicine to emotional states and everything in between. They are not more material than mental. Their first “evolute,” buddhi = “intellect”, awareness, knowledge - which the Gītā also calls Mahat, “Great One” - is cognitive, not material. The next tattva, ahaṃkāra, the “I-maker”, is psychological, emotional, rationalizing, and poorly understood if considered “matter.” The next tattva, manas, mind, is not treated as “material” in other Indian systems, and here indicates an empirically oriented mental function that “organ-izes” the senses, the cognitive outreach that engenders awareness of the world, of the sensorium. So the buddhi-indriyas open the buddhi-ahaṃkāra-manas continuum like a prism, into the five sensory capacities, the buddhi-indriyas, which are still cognitive rather than material. Only with the next set of tattvas, the sense-organs (ear, physical body [rūpa], eye, tongue, nose) have the tattvas concretized materially. From there comes awareness of sense-fields (sound, tactility, visibles, etc.), considered tanmātras, elementary factors of sense experience, and finally the material components that comprise them, the mahābhūtas (ākāśa [corresponding to sound<-ear<-hearing], wind, fire, water, and earth).

One might compare some of this to the arguments found in Nyāya, etc. for the layers of cognitive capacities, such that a self, over and above mind, etc., is considered ultimately responsible for ’knowing.’

Stepping back from the “history” of the model, it should be evident that this is a phenomenological account of how one progresses from potential cognate abilities to encountering the world. Note, for instance, the five mahābhūtas appear in reverse order from more standard treatments (more commonly earth, water, fire, wind, and when a fifth is included, ākāśa; some Indian systems eventually add consciousness as a sixth).

So just as in Buddhist accounts of pratītya-samutpāda there is a contrast between the pratiloma vs anuloma order — the former being the order of discovering, starting with death, #12, and working down to #1, ignorance, while the latter is the standard order for enumerating the 12 nidānas, from #1 ignorance up to #12, death — implying that thinking through things in reverse causal sequence is illuminating (according to Pali suttas Buddha discovers the links by asking “why is there death?”, and concluding “because there is birth, there is death,” he then asks, “why is there birth?”, interrogating each new link until ferreting out the full 12), the same applies to the 25 tattvas. The three guṇas are hardly self-evident without conceptual analysis and multilevel observations, but taking account of what we sense, and we process our sensations is ever present and open for contemplating. Start with earth = solidity, and work our way up to buddhi, the cognitive capacity to understand, to know, which reveals the guṇas and puruṣa in an understandable manner.

Understood this way, the 25 tattva model makes eminent sense in the order in which it is presented, and it is the spirit/matter superimposition which becomes counterintuitive. I hinted at this way of construing the model in an entry on Sāṅkhya for the Routledge Encyclopedia of International Philosophy (republished in Routledge’s Encyclopedia of Asian Philosophy) a draft version of which is online at http://www.acmuller.net/yogacara/schools/samkhya.html .

So what is your theory?

best,

Dan