According to MIL-HDBK-5, in the H1025 condition the compression yield of the plate is a little up on the bar (less than 5%). The plate is limited to four inches max thickness, whereas the bar is up to twelve inches. However, the tensile properties are essentially the same for the plate and bar forms.
As to why, one guess is that it's to do with the detail of the grain structure, but that's only a guess. These are S-basis properties, which are essentially what the producer feels can be a guaranteed minimum. Actual properties are almost always quite a bit higher. You might well find that the bar and plate properties overlap, depending on the batch.
As for 17-4PH vs 15-5PH, 17-4 was developed in the '50s and 15-5 is a mod of it developed in the '60s (I think). 15-5 has a better microstructure giving it better toughness, especially in the transverse direction. It also has better (i.e., slower) crack growth than 17-4 and is better when being hot worked (forged). MIL-HDBK lists 15-5 properties up to 12" and 17-4 up to 8". 17-4 remains a bit cheaper.
I'm not aware of any mechanical aspect where 15-5 underperforms compared with 17-4. We treat the corrosion resistance as equivalent, but that's getting away from my area. As airframers our main concerns are manufacturability and crack growth, which are our main reasons for selecting 15-5 above 17-4. For us the base material cost is a relatively small part of the final part cost, especially when total life-cycle costs are taken into account.