Nothing seems out of the ordinary for the weights. It's always harder to get younger respondents. More or less, if a subgroup is weighted up or down by more than a factor of two then that begins to raise some concerns.
A probability sample with roughly 1000 respondents has a margin of error at 50% is plus or minus 3% at the 95% level of confidence. At 15%, where most of the top candidates are sitting, the margin of error is closer to plus or minus 2% at the same level of confidence.
But, can a phone survey, whether conducted by live interviewer or by IVR really be considered a probability sample these days, where a probability survey means that every member of the target population has an equal chance of responding?
Edit: So, I would interpret the results more directionally. Currently, the top tier of candidates consists of Bailao/Matlow/Chow. Saunders is next, with Hunter/Bradford bringing up the tail. Chow seems to be more popular with older people, and is competing against Matlow for the Downtown. And the suburbs haven't coalesced behind a candidate.