Incidentally, 100 ohms is not a "best fit" common value -- that will violate Iz(min). Note that Iz(min) / Iz(max) is a factor of 10%, and 100 ohms is larger than 90.9 by about 10%. Prefer smaller resistances (and adjust power ratings upward as needed). 91 and 82 ohms should be easy enough to find, otherwise 100 || 1k can be used.
P_R is correct given the assumed resistance (100 ohms), but obviously it's a bit higher with the calculated resistance. By about 20%, or 2.7W. Because "10% more" (a ratio of 1.1) squared is about 1.2 ("20% more"); or exactly, 1.21.
Pzmax is simply Izmax * Vz or 1.1W.
For B, first, show the circuit. Consider the zener: what bias does it need*? Does it depend on the transistor characteristics? If so, which ones do we need to know to solve it? Should we begin selecting components immediately, or further consider the problem before looking up parts?
*Note that Vz varies with Iz, and diodes are rated at a given Iz. So, if we're varying Iz, Vz will vary some as well. For sake of simplicity, let's just assume that we can select a diode which has the correct Vz at whatever Iz we're using in the circuit. We can worry about that later (when it comes to component selection).
The transistor dissipation will be closely related to the voltage drop and load current; indeed for common components, it's fair to assume (because the error will be less than, say, 10%) that this is all of it, so we can immediately solve for the approximate power dissipation as simply
(Vin,max - Vout) * Iout,max
or about 1.5W.
Tim