Speedlink , lost a bucket of money (but was cross-subsidised by the trainload operation - especially MGR coal) - once the freight sector was split internally - the true losses were exposed , hence the end began before RFD came along. A great shame , but the wagon and loco productivity was poor (1 trip a week for many air braked vehicles with high local yard costs) - efforts were obviously made to regroup much of the better flows.
The train ferry element was the very worst earner when you put in the crippling costs of Dover port and the ferry.
You could just do "wagonload" freight using swap bodies or containers, then you could probably get considerably better utilisation of the expensive parts of the stock (ie. the part of it below the solebar) since any significant users would have access to a straddle crane or one of those big forklift things (they are suprisingly cheap these days).
Indeed, it appears that you could build a swap body that you could just leave in a siding by deploying its "legs" and pulling the wagon out from under it.
Then you would still have the benefits of being able to serve anyone with a siding (by leaving a short four wheel conflat with swap body/container on in the siding) but you could tranship from trip workings onto longer workings using a crane without time consuming shunting movements.
If you did go that route then a marshalling yard would just be a pair of parallel lines with straddle cranes and mainline workings could use CargoSprinters or
M250 series type EMUs. Indeed, freightliner workings could be similarily operated as I understand that they run the same rake of conflats every day even if it only has one container on it to avoid shunting moves.
You might not be able to reach commercially viable rates, but you might be able to cut costs (especially time consuming and labour intensive shunting moves) sufficiently to stand a good chance of atleast being self funding.
As for the trainferry issues... are the costs for sending a freight train through the Channel Tunnel now pretty much the marginal costs of doing so? As I understand it is far from full so they should jump at the extra income from more train paths surely?