I absolutely agree. My idea of having a symmetrical arrangement for speed and creativity is that there will be brilliantly creative people who need a lot of time, and amazingly fast people who have the creativity of a lettuce leaf.
In terms of +/-, because this is 2D, I would describe these as -3 + i and +3 - i.
Now, because everything is done per subject, you can be -3 -i in absolutely everything but basket weaving, where you might be +3 +i. Would you be a success or failure? Broad society would probably say failure, but I can absolutely guarantee you would have a very successful, highly profitable business and international acclaim in the art world. That sounds like a more interesting measure.
Ok, what about those who are truly doomed, negative in every aspect in every subject, learning slowly no matter what you do?
Well, one aspect of streaming is that nobody holds anyone else up, so such people aren't inhibited further by even slower people. In turn, they slow nobody up - a significant cause of classroom disruption that hurts those who are struggling even further. So these people certainly exist, but should fare a lot better.
With customization of style as well as speed, those "not getting it" because the presentation is wrong for them rather than any lack of ability should be running much closer to their natural speed. You have to introduce a third dimension to the streaming to tune the style better, but a mere 3 styles takes us from 15 streams to 45, and the total cost jumps from 2 billion to 6 billion. You might be able to do this - the law of diminishing returns won't kick in until classes (or buckets, if you like queue theory) cannot be kept full OR your research division (the actual end product as far as economists are concerned) have saturated the market, there just isn't any way to absorb the extra products.
Ok, but even when you have siphoned off all those doing badly for extraneous reasons, and got them where they can progress naturally, there will still be people who do badly. Some may even become politicians. The system I have outlined allows any one of these people to change gears. (In fact, it allows anyone to. Shifting down can help avoid burnout, shifting across can help avoid getting into a rut, shifting up can really stretch the mind.) So those who circumvent neural challenges (I'm one, so I know it can be done) can experiment. They can test any or all adjacent streams, without risk. They should be encouraged to, the differences in perspective may help solidify the person's methodology and lead to ever-increasing confidence and ability beyond the genetic baseline. Again, this is true of everyone.
Those who cannot beat the limitations should not, as often happens, be dumped by the system. Those who are slow should be allowed to continue schooling even to degree level or beyond, just at a pace they can manage. Those who have difficulties that mean they cannot learn certain skills at all, ever, at any pace, should nonetheless be encouraged to master what they can, as far as they can. I am stuck on ideas on how to help them further, suggestions are welcome. But what I am proposing is a definite improvement on what they have. Nobody gives up on them, as happens so often today, and because it's ok to suddenly "get it" at any time, nobody feels like they are necessarily marked for failure.
Added to which, you don't need to be a genius to be a lab assistant, and lab assistants are just as entitled to flashes of insight and inspiration. Such insights may lead to even better solutions for struggling minds. After all, rote memorization has little to do with ability. Indeed, very little memorization is needed. Those with poor memories but brilliant idea engines need a way to offload the part they will never be able to do, to nurture and turbocharge the things they are great at.
Ok, this pushes us into computer augmented learning. This isn't a new axis, since these are prosthetic aids that let you take advantage of your strengths. These aids should never be such that natural ability ever decays through under-use, but they should supplement those abilities so that people aren't disadvantaged by the irrelevant. Anyone can - and should - look up facts, because human memory isn't reliable. Memory is useful only as a temporary work buffer to learn skills. So if a computer provided that temporary work buffer, index and, indeed, knowledgebase, it does not reflect one whit on your skills or ability to utilize them.
There is currently no way to plug a memory expansion pack into the brain. It should be possible, though. Once it is, that kind of neurological disadvantage can be eliminated. Would it provide an unfair advantage? No. Because you can train a decision tree with facts, and train a rat brain to respond to stimuli to operate controls, a rat wired to an expert system can pass a traditional exam even though they could do nothing else. (It does prove that we are in a rat race, though.) Ergo, you need an exam that tests current understanding and usage of that understanding.
What is the purpose of an exam? In most schools, it is a barrier to the next level, where parrots and cyber-rats have all the best seats reserved. Here, the purpose is to see if a student is matched up correctly to a stream. There are no quantum leaps, no years in the classic sense, but an approximation to a continuum, enough gear ratios that you can slide smoothly around with minimal repetition and minimal synchronization issues.
Since there are no years, exams cannot be at the end of them. I picture exams as being an educational version of dynamic probes - if a bug in the process is detected, you test to see if the bug can be patched (extra help) or if there is a mismatch between student and environment (ie: stream). The student can then be matched better, re-learning only the bits not quite grasped.
The "final" exam, when the student transfers out of the system (regardless of when) would be intended to translate the level of ability in each subject into terms that can be understood outside. Thus, a person with the knowledge, proficiency, experience and demonstrated skills equal to a doctorate would have a doctorate. Time spent in the system would not matter, nor would a formal designation of being in a doctoral program. The exam would not be "for" something, in the ordinary sense, it would merely establish a level of competence, where the label is decided by what that level is.