Agreed there's no substitute for the 3x table, we should still do it when using VNAV and we HAVE to do it with FLCH. The standby glide slope is also a great tool: "It's showing a little high but that's ok because we're still on base" etc..
VNAV is great when we're flying a procedure: keep it on the path and it'll take you there: leave the LNAV path under vectors and it's suddenly useless. Or is it? We see it it all the time: "Fly heading xxx" - BAM! - FLCH, Speed brake out, extend the centerline. Stick with what you know.
Few people know or understand what the Vnav is doing, and fewer care to find out during the most critical phase of flight. If you've extended the centerline, Vnav is drawing a line upwards from a known point in space and a known speed/altitude constraint. The problem for us is that we're not flying directly to it, we're flying at an angle to intercept the LNAV/LOC, which makes our track miles unreliable. Here's the thing though: you ARE STILL going to that waypoint, which means that Vnav path WILL STILL funnel you to that waypoint at that altitude.
So if you get turned in early you can open the window, pop the brakes and follow the diamond down whilst doing your 3 times table as well. Opening the speed window and coming off the VNAV path into VNAV SPEED has exactly the same effect on the flight modes as FLCH. There is no difference between THR HOLD/SPD and THR HOLD/VNAV SPD: both hold the speed shown in the window and both control rate of descent with throttles and speedbrakes. It'll descend just as fast as FLCH, and once you get to flap 1 it'll even recapture the path and fly a nice neat CDA for you...