Depressingly plausible. :7
It’s not, though. There is no such alternative way.
If I do that; What I will really have done, is to have reduced both the horizontal and vertical angle, preserving none of them for the render, and subsequently stretched their limited content out over the larger span of the HMD, literally zooming in, and thus misaligning the game camera projection, and the view one, that it is supposed to correspond with.
…and if I were to just move the camera closer to the viewplane without changing the angles,¨then the view would still encompass all four lines and twenty rows, only rendered onto a “smaller” plane, because the camera sees the virtual 3D world behind the world plane, as projected onto it – not a 2D flatland that exists on the surface on the plane itself.
It is exceedingly kind of you to give the concept (…and by extension anyone who’d put it to use, whether on purpose, or out of ignorance), the benefit of rationalisation; Trying to understand why things end up certain ways, is always good practice; But zooming in is not something that is on the table to begin with; The whole point of increasing the FOV is that you do show that additional stuff to the sides, and if any VR developer thinks they just need to “fill in the whole screen with stuff”, with no concern for angular accuracy, then I have no words for their incompetence/negligence – they could do that by just writing #123456 to every pixel on screen. 
Getting hung up on the display panel and its aspect ratio, and getting it all mixed up with camera angles, is about the same fundamental complete and utter misconception of basic geometry (wilful or not), that has Pimax still to this day barf out the in every aspect inaccurate marketing claim that they have 200° field of view “diagonally”.
If anything; What I could do, would be to letterbox or pillarbox an incomplete render that does not fill out the entire visible FOV, so that it takes up the same FOV, on part of the screen in the large FOV headset, as it would on the whole screen in a smaller FOV one, which is pretty much exactly what happens when we choose small or normal FOV in PiTool.
(On a side note, btw: There was one dev (I seem to recall he was working with Unity), who made a reddit post some time back, about how he fixed pop-in (EDIT: …for his own title…) by adjusting camera distance, or something like that, and I must say that his approach sounded very much like a case of doing something-that-seems-to-work™, that circumvents a symptom by brute-force increasing the width of the culling frustum, rather than addressing the actual root of the problem.)
Well, regardless of how many objects we evaluate (and if we want to see more stuff, then that more stuff will need to be rendered - a bit of a truism :7), the number of pixels we need to render are the same, and AFAIK fragment shaders typically constitute the lion’s share of rendering load these days. :7
…and objects will of course have varying sizes and complexity, and be at aribitrary locations. I must admit that your visualisation example with the fixed block octree world evokes some Euclideon associations (shudder). 