The BF-2 is a decent flanger, but given the range of tricks that people expect a flanger to perform (generally a wider assortment than is expected of chorus or phaser pedals), it is not often the case that people find any single 3 or 4-knob floor-box flanger the be-all and end-all.
The most difficult trick to perform is the sort of ultra-wide sweep from very very "high" (shortest possible delay) to very very low (longest delay and well into chorus or double-tracking territory). One of the pivotal factors in being able to do that (especially the "high" part) is being able to provide a nice crisp clock pulse to the delay chip, even as it soars above frequencies around the 500khz mark. The MN3102 can provide high frequencies like that, but not unaided. And therein lies the shortcomings of the BF-2 and many comparable stompboxes; they use the MN3101 or 3102 clock generator/driver on its own, withuot assistance, so the pedal is generally limited in how high it can sweep.
Note that the capacity for wide sweep and tonal quality essentially have nothing to do with whether it is P-gate or N-gate. The primary difference between the MN3207 and the MN3007 is that the polarity of some things is inverted (i.e., they are not pin-for-pin plug-in compatible even though all the pins do the same "thing" or serve the same role), and the MN3207 is optimized for running at lower voltages, such that you can run it on a regulated 5V and the bias voltage remains suitable even as the battery starts to decline from 9v. Any claims or observations of this MN3007 flanger being better than that MN3207-based flanger stem not from the chip itself but from the remainder of the circuit that the delay chip is nested in.
Flangers based around the SAD-1024 have often enjoyed a somewhat better reputation than those based on the MN3007/3207 because the SAD-1024 handles very high clock rates nicely even in the absence of assistance. The MN3x07 chips could probably sound every bit as nice, but it would taker more parts and sightly higher production cost to do that. So, on a dollar-for-dollar basis, people have tended to be more appreciative of SAD1024-based flangers. Again, that remains, in my view, a function of the circuit support around it, and the extent to which its finer qualities are milked, rather than anything inherent to the chip.
The Ultraflanger does not use an MN3101 for the clock. It does, however, provide the "assist" spoken of earlier by means of the paralleled invertor sections that bjuffer the clock pulse going to the MN3007 and adding the sort of "oomph" that permits the clock to be nice and square even at very high frequencies.
So, if you can identify a circuit that provides a suitably buffered clock signal going to the MN3207, you should be able to squeeze some nice sounds out of it.
