Are you saying that you don't think it's appropriate for those who stands to lose more to pay more to prevent that loss? Again, back to the insurance analogy, do you think the person with the $10 million dollar home should pay the same price for their insurance as the person with a $100K home?
Insurance doesn't prevent loss or reduce its probability; it merely compensates bad fortune. Hence if the payout is unrelated to the value of the loss (which is usually the case in respect of social insurance), then how big the loss is to an individual is not related to the cost of insuring them and therefore is not a reason itself to charge them
any more, still less a progressive fraction more.
If you wished to refer to
prevention/protection, which is distinct from insurance, then the reduction in probability of loss
is worth more to someone wealthier, and if everyone could co-ordinate with each other and overcome free-riding, wealthier ones should
voluntarily agree to pay more for the protection. I don't think that is what you were arguing before. It is not clear that the value of the protection would be
progressively more, however--IE that it would be "worth" $10K to a $100K net-worth individual to have protection but worth
more than $100K to a $1M net-worth one to have it. So I don't think that gets to a justification for progressive tax either. (The diminishing "marginal utility of money" works both ways in that consideration BTW, and the outcome is not clear).
And to repeat a reply to someone else, the
cost of protecting a net-worth (or income; whatever) of $1M is very unlikely to be ten times the cost of protecting a net-worth of $100K, but more likely
less than ten times--so on cost-based reasoning the tax levied would be regressive, not progressive.
(I am not arguing that there is no justification for progressive tax--but neither cost-based nor value-of-loss based reasons provide it.)