Originally Posted by mm43
Not a good basis for PLL capture, bearing in mind that the ULB oscillator is most likely R/C with inherent phase shifting and possibly small frequency changes as the battery voltage drops.
However, I should point out that an uncertain frequency is one problem that can be aided with improved signal processing. In the traditional detector one would make the filter bandwidth wide enough to be sure to include the maximum drift of the signal, thereby including more noise. With DSP techniques, it's easier to simultaneously track multiple narrow bands to find one that has a signal. This is likely part of RR_NDB's point.