I think you are referring to a positive lightning discharge. If so, they are not particularly rare, and you've probably seen many of them if you've chased more than a couple of supercells. Positive strikes tend, in a normally-polarized thunderstorm, to originate near the anvil and hit the ground beyond the storm. In a basic sense, the electric structure of many thunderstorms can be thought of as a tripole, with the primary positive charge reservoir near the top of the storm, the primary negative charge reservoir in the low-mid region of the storm, and a smaller positive charge reservoir right near cloud base. Positive strikes tend to have much larger current / continuing current -- on the order of hundreds of thousands of amps! Both of these tendencies contribute to positive strikes being a significant fire initiator: a large continuing current + a strike away from the storm and precipitation + tinder-dry vegetation or structures = fire!. I'm not sure I'd give them the distinction of being called "superbolts", since they can be very common in some storms and environments. It is worth noting that some storms have an inverted polarity (i.e. primary positive charge reservoir is in the low-levels and primary negative charge reservoir is in the upper levels), in which case many of the strikes tend to be positive in nature. IIRC, storms on the high Plains have a much higher propensity to have this reversed polarity structure and tend to have a much higher proportion of positive strikes.
As for staccato or not lightning... I'm not sure I've read much about what would cause some lightning discharges to have many subsequent dart leaders vs. being "one and done". I know the storms of June 11-12 (I think) in KS seemed to have a lot of staccato bolts, with only one extremely bright flash that seemed to occur in a 'barrage' manner. I remember an HP supercell chase from 4/21/05 that had similar lightning patterns (a plethora of extremely bright but extremely short lightning discharges). I assume it's largely affected by the amount, density, 'wetness', and distribution of hail and graupel within the storm (as it is, the lightning flash rate tends to be correlated quite well with graupel mass <-10C and mass flux at -10 C, with overall charging largely a function of temperature and liquid water content). Perhaps the staccato vs. not is dependent upon the particular charge density and distribution... For example, a relatively broad charge region with low-moderate intensity (measured as coulombs per cubic meter or something like that) may not have enough remaining charge to produce a dart leader to give the subsequent flashes after the return stroke. So, when charge is transferred during the return stroke, the local electric field is reduced enough to inhibit dart leader formation. However, since the charge area is relatively large, you can have many one-hit wonders in a short period of time. On the other hand, a more compact but intense charge region may have a large magnitude of charge that would allow for subsequent flashes (many dart leaders). Just a guess.
