What are you using to model this response? Was this simulated anechoic?
This does not at all match with the models I have completed on this driver??
Bumping up the suggested sealed enclosure you listed by 2.0ft^3 and going vented changes things dramatically.
GO-NAD!:
A 10.0ft^3 net vented enclosure with the Tempest allows for anechoic response with the lower and upper -1db points at 20hz and 63hz respectively. The lower and upper -3db points occur at about 18hz and 83hz respectively. In other words very linear response. This is also assumed at 500 watts rms input power yielding 114.9db output over most of the entire range.
Internal Dimensions of 20.5" x 28.5" X 36" would allow enough displacement for the 3.25" X 18" X 39.5" slot vent and .5 ft^3 for bracing & the driver.
One key thing to keep in mind is that if you went this route you would not want to use more than 500 watts rms as the vent with the above enclosure and the stated power level is rated at 16m/s vent velocity. Increasing input power to the full rated 1000 watts rms would push vent velocity to 22m/s which would begin to compress some but would probably be fine for most music listening.
Now here is something pretty cool. The Dayton RSS390HF-4 (15") models nearly identical to the Tempest in the exact same enclosure dimensions with but at about 2db less output across the same range. The reason the output is lower is that you will not want to exceed 300 watts rms input power to keep xmax in check.
Either route will work very well for most rooms. For around $320.00 you could do two Dayton drivers vs. one Tempest but you don't have any additional "headroom" so to speak with them (assumes you have space for two) With either the Tempest or Dayton playing at 100db I doubt most anyone would ever be able to tell the difference.
The Tempest gives you a larger performance envelope but the cost is quite a bit more. Tough call...