While been stuck in DN ... can't help indugling myself this evening in contemplating some toy theory on SE, borrowing some computing metaphors. Being a fairly noob practicioner, obviously take them with as much salt as needed . Love to hear esp. advanced practicioners to call out any bullshit ;) 

Here it goes!

---- 
Basic Premises (inspired quite a bit by TMI)
One user process as "I". There’re many processes in the system, your vantage point in observing the world and self is always through a specific user-level process. But there’re many others .. eg. Processes for emotion, narrative self, certain memories, breathing etc. — they’re more system-level, or privileged processes. 

Inspection: This user process has certain inspection capabilities / reflection tendencies - e.g. you sense emotion states, you construct concepts to explain them etc. As you grow up, typically that user process gains more data and insights about other processes that reside in this system.

Reactivity and reaction chains: Different processes are all constantly running, and often compete for attention. A fundamental property of the system, is it’s highly reactive, and almost always result in a reaction chain (interesting Q if it’s a infinite series, or finite), from even the tinest sensory input.

They compete for attention by bidding on how interesting/urgent they are. Some are by design of much higher priority multiplier (eg self preservation, or “ego”).

Implications
Suffering due to limited user view, and imprecise blob reporting. Unfortunately, your user process has a limited view of the system. It’s blocked by default from many deeper layer of the system.

Suffering, a “subjective” view, perhaps could be understood here, as an evaluation by the user process upon the system as a whole.

But it’s not true global view. Nor even it’s accurate view - because it’s taking reporting from other high-level processes for their words (“The sky is falling!!” says emotion)

However, a more fundamental suffering here is subject/object alienation. The user process consider itself the subject, while everything else is object. This alienation cannot be easily resolved. 

What is the game of Insight (in particular in terms of PoI)?
The Insight meditation solution seem to be progressively peeling out layers of user process (the “I”) through an exercise by quickly attaching itself faster and faster to its present object (emotions, sensate, memories, etc.).  By realizing that it can attach to any other tiniest thing, then it realize it must be so lightweight that it’s almost nothing of solid existence! This reduces suffering, but won’t completely because the tiniest observer is still alienated, fundamentally.

Now, here’s a weird design assumption - the user process *must* attach to some object, at all times, to something.

Except, perhaps, at stream entry / Fruition?

SE, thus, could then be hypothesized as a moment by which the subject user process is unattached to anything (perhaps not even itself). That lends the tremendous freedom that it can be at home, w/o that urge to attach.

Progression of Insight
The user process gains progressively more system-level Insight by peeking into a few critical things: lower level layers (and data), process switching mechanisms, and inter-process communication. And thus form a more wholistic and precise global view.

For example: 

• By having direct access to granular data of e.g. emotions, and thus differentiating individual sensates that composes “anger”, deconstructs the concept and might reduce suffering. 
• By peeking into process switching mechanisms, it understands “Cause and effect”. A direct implication is it might now find the emotion chain overly reactive (and hyperbolic!), and give it a lower trustworthy scoring on its reporting.
• By peeking into inter-process communication between say “narrative self” and “emotions”, it observes direct co-relation between certain emotions and self evaluation, and the high variance in diff settings (“I’m the smartest guy ever lived”, “Perhaps I’m just not that good at math …”). Again, lower scoring.

Those are kinda common sense .. and happen as human development process, w/o meditation. The special effect about meditation is reduction of reactivity, and microscopic observation. Leading to the next topic ...

Concentration, as a function of reactivity and filtering
Usually concentration is understood as clearly and consistently perceiving the object.  In this context, it seems that the primary goal is instead optimizing for a low-reactivity environment that is conducive to Insight acquisition. (And perhaps why “calm abiding” is used in some Tibetan  tradition).

We need low reactivity, because it’s hard to peek into mechanics of the system, if processes are constantly spawning processes after processes (high CPU usage, chain reaction) and leave very little room for gaps for observation. Gaps make observing things like process  switching, inter-process communication, much easier.  (This is of course, generalized from the “cessation” idea - if cessation considered a full gap)

Further, concentration is about discernment / filtering …

• When focus on left nostril only in breathing, we’re actively filtering out all other data. 
• The other extreme would be entire sensate universe.

In good concentration, regardless of the scope of the view, they should similarly lead to low reactivity (i.e. calm) system state. However, the informatio acquired from those observations are necessarily different - microscope more on foreground individual sensations, and larger scope acquires more background information (e.g. emotion state) and meta-properties (texture, density, frequency, etc.) due to sheer load of data in a limited processing bandwidth.

Layered ring fences, Jhanas, and Stream Entry
In all modern operating system (e.g. iOS), there’re always many layers of protection - the kernel lives at the core, and user level process (e.g. Tinder) lives at the outmost layer. This is to protect the system integrity and prevent systems from reading / attacking each other. 

Jhanas, perhaps, could be viewed as a very low reactivity state, but also it gives you temporary privilege from the kernel to access to other processes.  

The phenomena that after SE, you gain easy access to Fruition, various states, and Jhanas, can then be understood as you’ve gained the trust for those accesses (for knowing how and not abusing them). 

A wild guess on non-dual tradition
This is a fascinating one. Instead of focusing on raw data and lower level invisible processses, non dual traditions seem to be doing sth., very different (and clever!).

Its premise is - if the problem is the user process’ desperate need to attach to things, can we rewrite simply the scope of subject/object, and consider them a whole?

e.g. if we take Mahamudra tradition, the pith instruction is sth. Like “Events as moving waves. Awarensss like unchanging ocean. View Events as reflection of Awareness. Awareness reflecting upon itself” .

A wild guess would be - that it is training the user process to rewrite its own definition of subject, and thus view anything (even "external" objects out in the world") it attached to presently as something of its own, and thus wholesome by definition - i.e. everything is exactly what it should be.  Seems like a really clever "hack"!  Exactly how it's done I have no idea though.