As a Brit, I'm very annoyed by this, they could have just got behind the Arduino (for example) but no that was apparently NIH (Not Invented Here) so they've spent a lot of public money on this.
How much though?
Thing is, it's not a complicated board. I actually work on bluetooth stuff for a living and knocking up a uC, accelerometer and a few other bits and bobs on a circuit board is very seriously not hard. It's sufficiently not hard that I did a custom one as a favour for someone recently. When I say it's really not hard, it's really not.
Now, I've been using the CC2541 mostly for historic reasons. I've been meaning to jump over the nRF51 (used by the micro bit) and head straight for the nRF54 because HAVE YOU SEEN IT? Bloody awesome.
That aside, the peanut gallery (i.e. me) would have done it differently. I'd have used a ST Micro electronics accelerometer not an Freescale one mostly because then I could use SPI which I prefer (I2C is squirly and flakey, and much lower power draw which is something I happen to care about) and because I've used various of the ST MEMS sensors, I know they work well, I'm familiar with the interface and so I'd not change unless there was a pressing reason.
Anyway the important thing was that I had an onion on my belt as was the style at the time.
But back to it.
First, the arduino is great, don't get me wrong.
However, this board is not an expensive one to develop. This way they get much more control and have all the peripherals they want which you'll note is a lot more out of the box. When the BBC do something, they also go in big, and this way lets them control the supply chain so there's not going to be any worry about sourcing, and sourcing exactly what they want when they want in whatever quantities they need.
And those are large. They do (or did?) give them away to every yeat 7 pupil. That's about 700,000 per year, which as far as I can tell is about 3x the number of arduinos sold per year. And vastly more than any I/O shield. In other words they are so much larger than the entire arduino effort that it would require a complete revamp of the Arduino supply chain at which point, again, it's easier and simpler to simply do it themselves.
It's also a much larger MCU than many of the arduinos and not AVR. The advantage of the latter is that sadly few people work in GCC-AVR compared to ARM, making it a slightly less reliable platform. And the larger MCU also allows for things like micropython.
Of course they could use an ARM based arduino, but now you're dealing with a slightly fragmented platform and of course shields for the I/O. So they'd have to specify precise models (then deal with the fallout when people find cheaper ones which don't work) for both the controller and the shields, none of the latter of which do exactly what they want.
So maybe they should have made their own shield, but by the time you're there you may as well shove an MCU on it and call it a day.
In other words given the size and scale of what they're doing it was likely a better choice to make their own than try to cobble everything together with arduinos.
I do a certain amount of school volunteering and this is another thing that fragments attention and class time. I would have preferred full-fat pupil owned Raspberry Pis for example, a little more expensive but an order of magnitude more capable.
They are a *lot* more expensive. The RPi doesn't come with much out of the box. You need a PSU (many USB ports don't give out enough juice), an SD card, and probably a screen and keyboard to get going. This device is much more arduino like. They're not really comparable at all.
Still the BBC is a law unto itself.
The BBC's a huge organisation with many many interests. It's never going to be perfect and it's never going to do precisely the same as you.