often the TOD does not change when you enter or change constraints on the descent path - because it is not affected by a new constraint, or lifting an old one doesn´t affect it.
Imagine a "idle descent" path that passes well under a restriction of "be at FL250 or below". Now clearing this restriction won´t affect the calculated path, because it never affected it in the first place!
The calculation of the VNAV PTH descent in the real 737 worked fine - but it was complex and had some potential pitfalls. Especially when winds were unknown (as they mostly are) or when ATC changed your lateral flightpath (as they almost always do) the calculation of the optimum path can get you in trouble. This is due to the fact that the FMS will plan the descent in an optimal way - without any conservative reserves. So if there is less headwind or the routing gets shortened you will immediately be "too high" to start your approach. This does not work well with the conservative and safety-oriented attitude that airline pilots have. Also the complexity and missing transparency of VNAV calculation make real pilots shy away from using it - don´t use what you can´t understand and doublecheck.
So speaking for my airline, the typical use of VNAV was seen in climb and cruise. In those portions of the flight VNAV works exactly like FL CHG, with the difference being that the target speed is calculated by the FMS.
Usage in descent was close to 0. We used the TOD and vertical deviation bar to "sanity check" our own descent calculation, but the VNAV PTH descent mode was barely ever used.