Is it feasible to introduce output key reuse checking?

Question

Reuse of output key P=H(rA)G+B (which effectively burns fund) could potentially occur if the same random r was used in multiple transactions to transfer fund to the same recipient (A,B), most likely due to the use of bad pseudorandom number generator. Is it feasible to introduce an additional checking step so that the second transfer reusing the same P simply gets bounced, pretty much like how key images are checked to prevent double-spend? I heard somewhere that output keys are stored in an array instead of a hash table; will this make the situation a lot harder than it is with key images?

P.S. This question arose in the context of the possibility of one-time receiving address, a nice bonus that provides privacy even outside the blockchain and can be easily implemented if this reuse checking is available.

Answer 1

Answered by smooth_xmr:

We looked at this but it would add additional permanent overhead and reduce scalability to prevent what is essentially burning coins, which is silly. Someone who wants to burn coins can do so other ways.