i.MX RT PAC shootout: comparing MCUs

I realized folks (including me!) may not have a basis to compare the i.MX RT MCUs evaluated in the PAC shootout. Here are the numbers of each MCU's peripherals, registers, register fields, and field enumerations as generated by raltool.

This should help us see why the imxrt1011 build is faster than the imxrt1189 build: there's less Rust code to compile.

I took these measurements by modifying raltool to figure each count after the combiner pass, just before generating Rust code. These counts should be about the same for what's generated by chiptool, since they'll share the same transformations.

It would probably be better to count each from the SVD sources, rather than after the chiptool & raltool transformations. That would be more representative of what svd2rust is working with, or what chiptool and raltool would work with before sanitizing the sources.