I am not sure that running a slasher is limited to validators - I think all you need is a Beacon node (see here for instance).
That said, the computational resources are intensive.
As of right now the proposer that includes the evidence of a slashable offence gets the whistleblower rewards, see:
whistleblower_index = proposer_index
(here).
But even if the whistleblower did receive the rewards, there are two other key problems.
Given the relatively low frequency of slashable offences committed (so far) and the low payout - looks like about 0.05 to 0.07 ETH (depending on the slashed validator’s effective balance) - running a slasher does not appear to be economically feasible. Note also that that reward is, in theory, meant to be split between the whistleblower and the proposer (even though current implementations give 100% of that reward to the proposer).
The other, arguably, more important problem is that reports of slashable offences can be stolen as soon as they are propagated (as far as I understand). Some implementations (like Prysm’s) allow you to disable the propagation. I guess the idea could be that if you run a slasher and validator - you would disable propagation and send the slashable offence to your own validator when it is chosen to propose a block.
Not sure what the solution is here. From what I can see (slash_validator - in Bellatrix specs - link above), the amount slashed is “burnt”- maybe taking a higher proportion of this and giving it to whistleblower would help. That said, I don’t think that’s enough on its own to make running a slasher economical.
Things seem to be working fine right now, but if this were something to be remedied in the future, maybe it would be worth considering minting new ETH to compensate folks running slashers. Again, given the low frequency of slashable offences to date, it would require a very small amount of ETH (on a network-wide basis) be minted to make running a slasher economical.
Perhaps a hybrid approach would work best - set a threshold for whistleblower rewards to make running a slasher economical. The whistleblower would receive a higher proportion of the slashed rewards (including a portion of the proportional slashing penalty), but if the total reward to the whistleblower is below the threshold then an additional amount of ETH would be minted to reach the threshold. This would ensure that the whistleblower is adequately compensated and, in the case of a large attack on the network, where the rewards are more than sufficient to meet the threshold, no new ETH would need to be minted.
