Adafruit M0 Adalogger as a cheap 10kS/s DAQ (?)

So I've been looking into how to use an Arduino as a DAQ, the topic has been brought up a number of times by various people this one I personally feel to be the most conclusive. We know from the Analogread reference that the arduino samples from the analog pin at roughly ~9kHz which is awesome. But when you try to store the data by printing it to the serial, sampling drops to about 1kHz at best if not lower. This is because the serial data rate of 115200 baud is just really slow in comparison.

Now with the new Cortex M0 chip and an onboard SD card that is using I2C (I think?) on the Adalogger, one should be able to write analogread to SD at I2C speeds which is faster than the analog sampling rate . Thus you should be able to actually record at the max analog sampling rate of ~9kHz and print it to the SD and then print the SD card contents to serial after the fact for analysis. Or am I missing something?

For my own personal application I'm looking to sample at around 10kHz for maybe 5 seconds but with a big enough SD card, one should be able to sample quite a bit in both resolution and time. Also, would the Cortex M0 have a different analog sample frequency than the 328P?

Pay attention that writing to the SD card doesn't become another bottleneck. I couldn't find a datasheet for the M0 Adalogger, dunno how many bits and conversion time applies.

So here is the data sheet for the uC IC If you go to pg 183 you see how you can modify the speed of the SPI protocol, as a function of the frequency oscillator. If the adalogger has a 48MHz crystal then the SPI protocol should be more than fast enough (?) So the SD shouldn't be a bottleneck. Additionally if you look at pg 292 from the same data sheet, the ADC has a maximum sampling rate of 15kS/s.

So from what I've read it should be possible to hit 10kS/s as long as the data is transferred to the SD card and not serial.

That was the incorrect data sheet, this is the correct one.