The main lesson learned, so far, is that attention to detail matters greatly, and every individual employee must be made responsible for it. Ultimately, each participant in the workflow needs to be responsible for ensuring and maintaining their own robust technology and remote connectivity in times of emergency, and if they can't connect before the storm they have to take ownership for fixing that. Not everyone will know how, but everyone can reach out for the expert assistance to make it happen, and they must.
It is too late to fix remote access issues when the storm surge comes in.

Of course you can't think of everything: in the midst of our troubles we ran afoul of another unlooked-for complication when our offshore technology team in India was evacuated due to a hurricane in Chennai. We are working through that now and pulling out plan C. But the bottom line is that we really do need those plans, A, B and C.

Be prepared, be prepared, be prepared.