Your first statement is accurate. However, the rest of your analysis holds true only if the speaker is operating in its linear range (not exceeding Xmax) during all levels of increased power applied, which is very rarely the case, if at all, when running out of, or near max volume. I'm a big proponent of doubling the speaker's cone area as you have a much better chance of getting a more linear SPL/Watt relationship, at least initially. Just adding more power to a given speaker rarely, if at all, gives you a linear return on power applied and only diminishes the tonal quality of the system with increasing power, IMO
Yes my comments are a bit simplified. As speakers are pushed closer to their limit, they experience power compression; meaning at some point you don't gain the full 3dB per doubling of power. Therefore calculated SPL is expected to exceed measured SPL.
Another issue: My understanding is a driver's mechanical power handling varies with how the cab is tuned. AFAIK, it's not unusual for mechanical power handling can be significantly lower than the driver's RMS power rating. Unfortunately some companies seem to rate their bass cabs using the driver's RMS power rating and provide no info on mechanical limits. Off the top of my head, Mesa and Acme are the only two companies that specifically address mechanical power handling in their specs.
The right choice can depend on the situation and how close you are operating to the limits of your gear. Within a certain operating range, increasing power results in superior sound to my ears. Above this range, it's generally better to add a matching cab.