FYI, the spec sheet for RMG SM 468 shows a 2.0 dB difference in dynamic range at 30 ips vs 15 ips via both IEC and CCIR bias noise measurement standards. For SM 911, they show 3.5 to 4.0 dB difference in dynamic range. I've never noticed a difference bigger than that.
Also, in the real world, frequency response will probably not actually double, since the 15 ips response could easily be pretty good out to 25K (like my little Otari 8-track) and the overall system (tape and electronics) may still start to roll off at 30K at 30 ips (and most all your mikes start to roll off big time by 20K if not below). Mikes that don't roll off until much higher tend to be noisy, because they have to be so small and the self-noise becomes a problem relative to the signal produced by the capsule.
As you say, you do typically lose about an octave off the bottom, though newer Studers and the Ampex ATRs are generally able to keep even the 30 ips response flat down into the 20 to 30 Hz region.
Cheers,
Otto