Even if this were true, that levying something like a $50 tax per tonne would result in
greater incentive to pollute (because we've replaced a powerful social disincentive with a less powerful financial one), it would still very likely reduce co2 emissions.
Cost of energy is a significant cost in most mass-production processes (particularly in manufacturing). So, prior to the imposition of a carbon tax, when faced with the choice of purchasing one of two identical widgets, a 'co2 indifferent' consumer would choose the cheaper one. All other things equal, if widget manufacturer A can externalise some of their production costs (by polluting for free) and widget manufacturer B can't (or won't), widget B will likely have a higher price, even though their total social cost of production is lower. So there is more incentive to be a polluter.
The situation is reversed if a tax is levied on co2 pollution, with the rate set to internalise the cost of all external harm. Now the 'co2 indifferent' consumer will purchase co2 free widget B, because it now has the relatively lower price (prices are now more reflective of total social costs of production because the global cost of polluting has now been internalised).
I guess the other approach, that sidesteps this whole argument, would be to implement a 'cap and trade' scheme (my personal preference). The government auctions or allocates, say, 100,000 one-tonne permits: each permit allows a business to legally emit 1 tonne of co2. If a business needs to emit more co2, and what they produce is more valuable relative to what some other permit holder produces, then they can buy additional permits from this other 'less productive' business (and all businesses like it).
Under this latter scheme there's a hard cap on emissions. So people can feel as entitled to pollute as they like; they'll still be breaking the law if they do so without owning sufficient 'pollution permits'.
Might the latter sort of approach (cap and trade) address your concerns?