For future EV design, I have tentatively chosen 0.157 kWh per kg. for energy density of the battery pack. Am I aiming too low for a vehicle likely to reach production in five years? Or, is my number too high?
Your initial figure of 157Wh/kg is representative of a first generation Nissan Leaf, but that's been improved on by Nissan (using automotive-specific cells) and other vehicle manufacturers using commercial format cells (mostly 18650 format).
Also, if you're after a general figure, be aware that it'd easier to get good packaging density on a big pack than a small one (there are always overheads in the form of enclosure, contactors, cooling, etc that all packs need to carry.
For this reason, I'd suggest you use a y=mx+c linear relationship.
My estimate of this for automotive packs between 5 and 50kWh, for current cell technology, would be:
Pack weight (kg) = 4 * pack capacity (kWh) + 80
(the coefficients are nice round numbers, because the answer is going to be approximate)