If only it were made more clear that there very likely exists a point in time beyond which the costs incurred from having to deal with any particular observance or non-observance of DST are dwarfed by the costs incurred from uncertainty and duplication of effort in the lead-up to and in the wake of last-minute policy changes.

I can't claim knowledge of the relative size of potential economic impacts... but this close to a proposed change (~a week out), it's almost certainly not a simple matter of "we can either change the clocks or not, and that will either incur a cost or not", but rather the (apparently) more complex matter of "costs are currently being incurred already because no decision has been finalized, and costs will tend to decrease once a decision has been finalized either way (although perhaps by unequal amounts)".

It turns out that, in our interconnected world, people, corporations, and indeed entire industries like to be able to plan timings reliably more than a handful of days ahead.  While folks are generally able to collectively deal with uncertainty like this, preparing workarounds and creating backup plans, we're well past the point where the total human effort involved will be minimal... and since the preparations currently must involve at least one potential outcome that won't ultimately come to fruition, those costs will likely only grow until people are finally able to focus those resources solely on the outcome that will actually happen.

</preaching_to_the_choir>