Exactly what I was thinking. These are hardened steel on steel threads. 43.5 foot-pounds makes not the slightest sense at all.
Not sure of the thread size on this application, but the usual specs for, say, 16mm threads (that's on the low end; countershaft threads are almost definitely bigger) are usually between 180 and 247 foot-pounds depending on fastener grade you're assuming.
https://www.boltdepot.com/fastener-information/bolts/Metric-Recommended-Torque.aspx (No idea why this chart is in NM.)
Maximums can be even higher, depending on grade, desired bolt stretch, etc.:
https://www.engineeringtoolbox.com/metric-bolts-maximum-torque-d_2054.html
NM to foot-pound converter:
https://www.theunitconverter.com/newton-meter-to-foot-pound-conversion/
In any case, it's quite clear that 43 foot-pounds isn't even near the ballpark.
And yes, I'm sorry to say that the revered FSMs for vintage and modern Suzukis are chock full of egregious errors like this and should never be followed blindly; the many errors in FSM torque specs especially have caused no end of angst and misery... always engage brain before arm.
If the issue continues after proper torquing, the PO would do well to remove the sprocket if possible and examine the splines for excessive wear that might be allowing the sprocket to "work" back and forth.
