Since virtually no one has been able to obtain a RasPi two months after "launch", I suppose a case could be made that it"s a failure, at least for the 2012/2013 academic year. Even if it had shipped in March, the state of the OSes alone is below what it needs to be for the general academic environment, much less education-oriented applications, documentation, training, etc.
March is too late for most school systems to plan their budgets to buy any, much less many, RasPi boards. Those that do make it into classrooms will be via early-adopter geek teachers and parents who are desperate to do something, anything, to improve academic opportunities. So, the coming academic year will be a beta (and to some degree, alpha) test period for the large number of moving parts (boards, cases, peripherals, OSes, educational applications, programming language configurations, documentation, etc., etc., etc.) that make up the RasPi ecosystem.
It"s going to take much more than a year just to get non-geek teachers trained in how to integrate the RasPi into their curricula that are already overburdened with exam-oriented requirements. It would be interesting to see if anyone can dredge up the development and deployment schedules and budgets for the BBC Micro, etc. I"ll bet you any amount of silver and gold that it didn"t happen in six months – it was probably more like six years before every kid already in school when those computers were introduced had been through the official program for at least a year. So, it will be at least several years before success or failure can be ascribed to the RasPi.
As for Linux and Python vs. anything else, it really doesn"t matter. The LXDE desktop (or whatever else may be used with OSes other than Debian) is more than adequate for the purposes of not scaring novice kids off, while offering enough hints of wonderful things under the covers to keep the brave and foolish experimenters delighted for years. I think it would be a better use of the OP"s time to adapt the tools being provided with the RasPi than to try to cobble together yet-another OS (just what we need, another "standard").
It should be noted that it's very likely that the RasPi will primarily be introduced into secondary schools during the first, year, or so, and if anyone is really smart, they'll challenge the students to do all of the heavy lifting for developing and documenting the plethora of education-oriented material that's desperately needed yesterday just to get the teachers up-to-speed. By the time the RasPi is really ready to be introduced into elementary schools (beyond the natural geeks, who will have RasPi boards bestowed on them by kind parents, grandparents, uncles, aunts, et al), some of those high school students will be graduating from colleges and universities as teachers themselves and will form the real bow-wave who will help complete the flood of classroom battle-hardened RasPi systems into schools. Fortunately, that will be right about when the many Boomer teachers and administrators will start retiring in large numbers, who will have absolutely no interest in learning about anything new, other than the cost of beer and electricity in Mallorca, Menorca, and Benidorm.
If you really feel the need to sugar-coat Linux, why not just create a set of commands with whatever you feel are more meaningful names and alias/link them to the "real" commands? Before you get too far down that road, people with similar ideas did that upwards of 30+ years ago, mapping Unix commands to their CP/M and PC/MS-DOS equivalents, like ls to dir, cp to copy, rm to delete, etc. I"ll leave it as an exercise for the reader to go find them – they should all still work perfectly well.
BTW, renaming rm so that rm -r /* (remove all files starting at root and proceeding recursively into all directories) no longer works would be a highly-recommended first step (assuming you were foolish enough to allow a novice user to be root). Obviously, enforcing use of sudo and not allowing users to ever become root explicitly is probably the second thing I would do, but, you might think something else is more important.
I would be curious to see what you think would be a better file system and security model than that provided in Linux. Microsloth thought NT was just the greatest thing since sliced bread until folks like me wandered into trade shows where they were exhibiting and proceeded to break into all of their demo machines on the show floor. We put embarrassing things on the screens that even the poor tech support weenies they brought along couldn"t figure out how to get rid of before prospective customers saw them. Oh, sure, we did it to Unix vendors, too, where the sales people were allowed to set the passwords, etc. However, the more hard-core companies posed much more of a challenge – then, it was just a matter of social engineering to get the sales folks to slip up, especially when we managed to get them to invite us to their parties where the spirits flowed freely (and, yes, it still works to this day ).