Clark's fortunes plummeted in tandem because in Michael Moore's immortal words, Clark is Dean 2.0. The Kerry/Edwards "hair surge" hurt the general-lobbyist who didn't go to Iowa nearly as bad as the doctor-governor who didn't go to Nam. Now the hole piggy-back scenario is screwed.
If you look at my recent poll, a subset of the Dean supporters are those who think Clark will beat out Kerry for second place. I only had one Clark victory option and nearly no one has voted for it (probably just being realistic), so I suspect the Clark supporters are voting KDCE, KECD KDCL, or DKCE, and DKCL.
My point is, those are the margins of Dean's slight edge currently in the poll (Dean 48, Kerry 46), because almost all the Kerry votes are clustered on the KDEC finish, whereas Dean's votes are mainly split between two alternate scenarios: DKEC and DKCE, with 14% supporting the Clark comes in third scenario. If those people (or their equivalents in NH) vote for Dean, Dean beats Kerry in my poll. This preserves Dean as Clark's foil and blunts the success of Kerry and Edwards.