If XCP is used, a solution to control the rate of deflation will need to be found, as unlike Ether, no additional XCP will be created. That, or XCP consumed could be frozen for some period of time, imposing an opportunity cost.
If “XCPC” is used, the rate of generation and caps on the max number of XCPC in existence at any one time will need to be discussed and established.
--- XCPC token would be rather weird - yet another token, periodically freely airdropped to XCP owners, essentially at no cost.
To execute a smart contract you can
1. buy some XCP,
2. get free gas (XCPC),
3. use the gas and sell your XCP.
--- Freezing instead of burning? This doesn't solve the problem. If many people freeze their XCP, it leads to the same distortion. The rate of freezing would have to be decreasing.
--- Decrease the rate of deflation - maybe the best option?
Re: freezing, I'm not sure the amount of XCP frozen at any one time would be material to the total amount of XCP in existence. If that is not the case, then some "scale down" coefficient can be introduced that will reduce the amount of coins frozen for a given smart contract execution as the total # of coins frozen at any one time increases. Schemes like that do add complexity, however, and I'd like to avoid them if it's not deemed totally necessary, at least for a first pass.
Freezing as a solution entirely depends on parameters. Especially time. Short term freezing is similar to XCPC solution. Long term freezing is similar to burning. Actually burning is a special case of freezing (infinity freezing).
If not really long-term, freezing basically means a near zero cost of the execution of smart contracts. Most XCP holders hold long term anyway.
First, the idea of zero cost goes against the principle of "gas". Second, XCP token would not be much useful in such a situation.