It's important to note that "Rule 1" is considered to apply automatically when referencing ASME Y14.5, but not necessarily so if using the ISO version.
Nonetheless, the concept of Rule one extends the size tolerance to be inclusive of the FORM tolerance. In other words, if the pin was as it Maximum Material Boundary (MMB) - it's largest size - it would have to be perfectly straight and be able to fit down a perfectly sized "thru-hole" of that same diameter.
As it departs from it's largest size, it can start to bend, waist, taper, barrel etc. WITHIN it's limits of size. HOWEVER, its local size at any point along its length can never be smaller than the Least Material Boundary (LMB) - smallest diameter size - tolerance. Imagine always being able to pull a perfect sphere of LMB diameter along the axis of the part and never having it be larger or "protrude through" the part.
With this concept understood, we can now consider how your question regarding Cylindricity comes into play. Cylindricity essentially controls the outside "shape" and "form" of the part similarly to how Rule 1 does, except that it is used to FURTHER REFINE that tolerance. Say, for example, you had a pin with a rather large tolerance on its diameter. You might be willing to accept the pin at any size, but you need it to be relatively straight and round. In this case, Rule 1 - although it also controls the "roundness" and "straightness" inherently, might actually allow TOO much variation in the part. To fix that, you'd apply an additional Cylindricity tolerance. Now that diameter of the part can fluctuate within it's limit of size, but would have to be consistent within the limits of its Cylindricity tolerance along it's length.
In summary, the limits of size - in this case 33g6 - are often sufficient enough to control the form of the object provided the requirements of Rule 1 are also implied. Where even tighter control is needed, you might add an additional Cylindricity tolerance.