I'm not sure there's any such thing as a single best system, but I do go looking for models, and have come to appreciate Norway's. This is a follow-up to my previous post in which I focused on statistical methods by which we might approach this question.
Average student attainment Norway ranks high here, behind only Australia and New Zealand, and is number one in mean years of schooling already achieved, at 12.6, for students aged 25-34, according to UNESCO. (For Americans who might be wondering, we rank third, at 12.4, with New Zealand just ahead at 12.5.) The UN's Institute for Statistics calculates that currently enrolled Norwegian students are expected to push this way up, to 17.3, and this rate of increase is something that should give Americans pause, as the generation of American children in school is projected to no longer rank in the top dozen if current trends continue.
Average student achievement Norway is a thoroughly middling country in its PISA scores, having, like the United States, a most recent median PISA score of 500. Norway's relative scores have been rising recently, unlike many other Western countries, perhaps in response to reforms to more highly value knowledge. (To give examples of how anti-competitive Norway's approach to education has been in the past, I'll point out that Norwegian children normally receive no grades at all during primary school, and even their secondary school leaving exams, typically taken at age 19, only require the sitting of one examination, in an advanced subject that the student will have chosen well in advance.)
Approximate knowledge and skills of the average young school leaver Multiplying attainment by achievement scores, Norway still holds on to the fourth position, behind New Zealand, Australia, and Korea. The USA, for those interested, remains out of the top dozen, behind also, in alphabetical order, Canada, Estonia, Finland, Germany, Iceland, Ireland, Japan, and the Netherlands.
Employability of non-college-bound students Individual transition plans are very important for special needs students leaving formal education for the "real world". Norway has worked hard in recent years to upgrade these, which appear to be relatively effective: 45% of former special needs students are employed in Norway, which is almost the highest figure internationally (Switzerland's is 52%). Norway has also been moving towards apprenticeships, a German best practice, to help its students who are unenthusiastic about higher education, and has also started return-to-training initiatives that have proved relatively effective. Possibly as a result of these various reforms (the existence of plenty of North Sea petroleum doesn't hurt), Norway's current unemployment rate of 3.4% is almost the lowest in the Western world. Wouldn't Barack Obama like to have a number like that right now?
High-achieving universities Norway is undistinguished here, perhaps because its secondary school leavers have not been rigorously prepared (I think this the greatest weakness of the Norwegian system; commentators recently estimated that, on PISA, high-achieving Norwegian math students were three years behind their counterparts in Shanghai). The top-ranked Norwegian university appears to be the University of Oslo, #75 in ARWU and #100 according to QS; THES has it ranked below the University of Bergen, which is only #135. (Again, I don't trust these numbers to be all that precise, no offence intended towards their creators, but I do think they are generally indicative of trends.)
On balance, what are the strengths of the Norwegian system, and what are the trade-offs? Perhaps a key strength to Norwegian and Nordic education in general is not to be in a rush with children. The Norwegians require ten years of comprehensive education under their slogan "One School for All", from ages 6-16 (they start a year earlier than some of their more admired Nordic neighbors--researchers have concluded that starting school too late correlates with lower test scores); this is consistent with their goal of "a high general level of education in the entire population", which I assert (a) they are achieving more than any other nation and (b) is a worthwhile goal to pursue, rather than the highest test scores. This may be a trade-off, since the fastest way to raise test scores is to kick out (to where?) all the low scorers, and since it is difficult for any impatient would-be miracle worker of a politician or superintendent to claim credit for miraculous rises in attainment the way they can for miraculous rises in test scores.
Also, the Norwegian system intelligently provides a place for everyone, for all children and adults, with their various talents and ability levels, in accordance with a vision of a society "where citizens master the art of living together"; this may be contrasted with a society and government of the people by the highly educated for the well placed vested interests, which is what I fear American society has become. The only trade-off I see to pursuing such a vision is that it will upset entrenched interests and also those who love America and want to patriotically believe, as a default position, that our customary way of doing things is automatically the right one, in spite of considerable evidence to the contrary. It's hard to change systems fundamentally, and requires a lot of patience and persuasion, as I am attempting to use.
I also commend Norwegians and others in northern Europe (including, and especially, the new government of the United Kingdom) for their legislation allowing the public funding of privately managed schools, since both of which factors, as independent variables, are associated worldwide with greater student success.
And this brings me to the chief way to proceed with this kind of information, easily available via the knowledge network of the Information Age. I think we should (1) start with a good general system design, like Norway's, and then (2) make it better by seeking best practices as discovered by researchers investigating independent variables, (3) remembering always to check the internal consistency of these attempted improvements, since in reality variables do not operate independently, but are embedded in systems in the real world lives of students and teachers.