I disagree, respectfully but most vehemently. In 1912, Bryan was already a three time loser and would have been viewed as a last resort retread at best. Marshall was a relative nonentity from Indiana who never got more than favorite son consideration for the big job. Clark...he might have had a shot in a conventional race, but not when facing TR.
In a straight fight, that might well be so, but not in a three-cornered one like 1912. All that any of these men need do is get the normal Democratic vote out, and the Republican split ensures that he will coast home - shopworn nonentity or not. Whatever label TR is currently using, he's still a Republican as far as the average Democrat is concerned, so his inroads into the Democratic vote are likely to be relatively modest - certainly no greater than in 1904, and probably less.
Bryan may frighten off some of the more conservative wing of the Party, but the beneficiary from that is likely to be Taft rather than TR - conservatives have no reason to switch from one dangerous radical to another. Of course TR could in theory run a more conservative campaign, but if he were being as cool and calculating as that, he wouldn't be on the ballot in the first place, but would be sitting things out, waiting for Taft to self-destruct, and making his plans for 1916.
As for Clark, he won more Primaries than Wilson, and scored extremely well in Progressive-leaning states like Illinois, where he beat Wilson approx two to one, and California where it was almost three to one. These results suggest he will run at least as strong as Wilson among progressive Dems, and might even do slightly better - perhaps winning California, which Wilson lost. This might be offset by a loss or two in New England, and maybe New Jersey, but even that is uncertain as I understand he won the Massachusetts primary by a comfortable margin.
Marshall, I agree, is the weakest of the three, and perhaps the least likely to be chosen (my money would be on Clark) but I can't see it mattering. He's an acceptable, middle of the road Democrat, whom all wings of the party can live with if they can't enthuse for. All he has to do is equal Alton B Parker's 1904 performance (the worst in the Party's history at that date) to give himself a popular lead over TR of between five[1] and ten percentage points - enough for an Electoral College walkover.
Marshall, in fact, would be the Democratic equivalent of Warren G Harding - uninspiring but uncontroversial, and more than adequate against a bitterly divided opposition. Harding, you'll remember, was plucked out of nowhere in preference to half a dozen more impressive figures - and still won easily against a Democratic Party hopelessly unpopular and at odds with itself, much as the GOP had been in 1912. That would have been Marshall's position - even without factoring in the sympathy vote he would probably get for having to replace Wilson in such circumstances.
[1] Five on the very unlikely assumption that
all of Marshall's "losses" translate into TR gains. In fact, they would probably split between him and Eugene Debs, a former Democrat who would find it easier to win Democratic defectors. If
none of Marshall's losses go to TR (equally unlikely of course) his lead increases to around ten points.