Actually your last plot looks very good, the curves seems to be MORE parallel than the curves from the reverence strip. If using Flexicolor I had mostly times of 3:25 min, somewhere in the processing manual is written that times can be adapted within certain ranges to maintain an optimal result for rotary procesing. You don’t have to stick at 3:15, a drum development is simply different from a continuous industrial development machine.
Does your sidekick do a wash step after development? If so, skip that and go directly to the bleach step, or preferable to a stop bath prior bleaching.
And this may be the most important part; a prewash is originally not intended in C41, this will lead to (slightly) different densitometer readings than with films processed without prewash. This you can read somewhere in the process manual. The trick is that you normally do not see the difference in the print, and a proper tempering in home processing is quite a struggle without prewash. So it’s a bit up to you weather you want chase behind process results of a different development machinery (continuous cine type machines or minilabs) or accept minor differences and try a print first, to get an overview of the “visible quality” of your Process. Again, your curves look more parallel than the reference…