[WorldMaker]

Many of the least efficient vehicles like the EV Hummer the steel frame is still heavier than the battery.

The efficiency game as EVs scale up will always try to get the highest range out of the fewest batteries/battery cells, so cars don't necessarily "only get heavier" because weight impacts efficiency (as does drag coefficients and overall surface area). Lower weight cars get more battery range from the same size battery.

Also, while battery cells sometimes get heavier as the charge capacity gets denser, it is not a straightforward relationship. (Look at phone batteries over the last couple of decades: they store more charge than ever but are also effectively "lighter" in weight than ever.)

[WorldMaker]

Relatedly, because it is fun to speculate about other ways that battery weight may shrink in the next few decades though we don't have enough data yet: right now we're in a "range race" where manufacturers in the US, especially, are all focused on getting the highest possible range numbers (and thus more battery cells to cover it) that they can and meeting/exceeding the 300 miles per full charge "goal post" to fight Americans' current EV range anxieties.

That "goal post" itself may shift, or at least open up to be more of a spectrum of consumer choice. As more Americans get more home charging infrastructure they may grow much more comfortable with smaller range batteries. The early Nissan Leafs have already shown that some home owners can successfully live on 100 miles per full charge or less batteries (even some road trips with current charging infrastructure). It may make sense for car manufacturers to compete on EV form factors that intentionally include smaller batteries and overall are much lighter-weight cars, if for no other reason than it is a way to drive costs down towards extremely cheap low end models.

That's not entirely imaginary speculation either, because Asian markets with faster EV adoption curves than the US average already have some really interesting competition on light-weight, low-range, extremely cheap "urban" EVs. There's probably plenty of market potential in the US for sub-$10k low weight, low range EVs as at least secondary cars in households, and in some cases (some cities) would even already meet many primary car needs. It's just probably going to take a lot more infrastructure, marketing, and consumer convincing in the US market past current "range anxiety" over-concern before we see such cars common in the US, though.