PID_CALIBRATE advice test max power after calibration

[nefelim4ag]

Right now, some heaters can produce suboptimal pid values after calibration, they are suboptimal because they produce large oscillations during standby or printing. Often, it looks like I and D are large, and decreasing them helps.

That happens, that the calibration of the heater when the time of heating is close (equal) to the time of cooling produces less aggressive PID values.
Those values in the end should be less annoying and less often produce several degrees of oscillations, which can block the machine on M109 indefinitely.

This PR adds a new param to the PID_CALIBRATE, MAX_POWER, which allows adjusting power during the calibration.
On top, here is a small analysis section, which advises/guess MAX_POWER value to use next.
The point is, reduction might be necessary during calibration, but it is not necessary to reduce power for real printing.

*Based on Kevin's idea, I've only implemented it and gathered some testing data before

Thanks,
-Timofey

---

Testing:

cd ~/klipper
git fetch origin pull/7209/head
git checkout FETCH_HEAD
sudo systemctl restart klipper

PID_CALIBRATE HEATER=extruder TARGET=235
// PID parameters: pid_Kp=35.189 pid_Ki=9.022 pid_Kd=34.310

Graph with default PIDs at steady state:

With recommended power:

PID_CALIBRATE HEATER=extruder TARGET=235 MAX_POWER=0.599
// PID parameters: pid_Kp=27.901 pid_Ki=6.643 pid_Kd=29.297

So, for me, there is no significant difference.

---

I did try to calibrate my bed with decreased power (18%), but it seems to make I too low (0.064), so it has too slow heating and overshoots more than I prefer it to.

[aka13-404]

PID changes without restarting WHEN