ASUS Zenbook UX430UAR touchpad sleeps (seemingly not a USB autosuspend issue)

**The System** This is a brand-new ASUS Zenbook UX430UAR dual booting the pre-installed Windows 10 and a minimal Debian, uname -a: Linux myhostname 4.9.0-7-amd64 #1 SMP Debian 4.9.110-3+deb9u2 (2018-08-13) x86_64 GNU/Linux **The Problem** When the touchpad (motion and tap) has been idle for about one second (definitely less than two), the next tap is ignored. The motion response also lags noticeably after being idle, and it seems that the longer I am idle, the more pronounced it is. The motion lag disappears after a moment of use. **Why It's A Problem** This is a bad situation because in order to safely execute a single click by tapping, you must wait at least one second, then double-tap. You can't just double-tap as a rule, since if there was a motion event within one second of the double-tap, it will be registered as two clicks. Or you could just single tap as a rule, and hope for visual feedback to tell you whether the tap worked or not. If it takes you more than one second to evaluate the visual feedback, the cycle will be repeated! **What I've Tried**

• I've scoured about two dozen pages for ideas
• "libinput Disable While Typing Enabled" was 1 by default, but I have set it 0 via xorg.conf.d and the setting does persist across reboots
• I tried various things to disable USB autosuspend, including disabling autosuspend on all devices using the "usbcore.autosuspend=-1" and/or "usbcore.autosuspend_delay_ms" kernel boot options -- these seem to take effect, to judge by the /sys/bus/usb/devices/*/power/autosuspend files, after rebooting. (I've removed the boot option now.)
• I ensured that none of the following are installed, since they are known to interfere with manual power management configuration: laptop-mode, powertop, pm-utils, tlp
• I tried installing the xserver-xorg-input-synaptics but libinput continued to be used (and I dare not remove libinput lest I be left without a keyboard)
• I tried two different window managers (my own hack of twm that I normally use, as well as stock fluxbox)

Unfortunately nothing I've tried has influenced this touchpad misbehaviour in any perceptible way. Is there some device file I could echo to in a once-per-second loop to keep the touchpad from being idle, yet without triggering actual events on the desktop? (This is probably what I'll try next, as I await answers or comments.) Can anyone suggest something else to test or try, so that the touchpad is as responsive after being idle as it is in the heat of action? These are popular notebooks and a pretty good value, so a solution to this stands to benefit many.