Not really. Many of us have suffered autopilot glitches. It's not really less workload during an approach, because there is much less margin for error, particularly on a precision or LPV approach with low minima. You have to catch a pitch glitch instantly during an approach, or you may have some real problems. Hand flying does not suffer the same type of glitches. Particularly for two axis coupled approaches with VNAV and multiple stepdowns, there is a lot of system management to do to keep the VNAV from doing something unexpected. Things like, GFC700s stop giving VNAV guidance during a course reversal or while tracking a VOR or LOC (without GS), but give unofficial VNAV guidance on an LNAV approach. The ALT SEL has to be set lower than the next stepdown until the GP/GS is intercepted or it will not descend. And so on. There are a lot of button presses and mode checks to be done in a short time, and getting one wrong can have a lot of consequence. Even if it's working 100% correctly.
And the couple of "wrong" GS intercepts and uncommanded trim windoffs I've had (in both directions now) have made the automation considerably higher workload than hand flying. Ever try to hold a GS when the autopilot has wound trim all the way down in a 182? Yes, you can (and should) turn it off right away, but it is still FAR out of trim.