The SC House poll is probably the most reliable, being based on a sample of 2400 likely voters with only a two percent margin of error. The PPP poll is based on a sample of 897 likely voters with a 3.3% margin of error. All of the others are samples of between 375 and 502 likely voters, with margins of error between four and five percent.
All of them project Trump winning handily, with Cruz and Rubio in a tight race for second. Thereafter they differ greatly, and in important ways.
The SC House poll shows a virtual three-way tie for second place with Cruz, Rubio and Bush all within the margin of error. Such a result would do little to accomplish what the "sane lane" needs to accomplish, and as quickly as possible: to have Rubio, Bush or Kasich (most likely Rubio) emerge as its consensus candidate, with the other two withdrawing and endorsing him. If anything like the SC House poll is what the returns look like on Saturday night, we'll have to wait until Super Tuesday on March 1 for any movement toward that happening. And if the process takes longer than that, it might be too late. The battle might end up coming down to Trump and Cruz, with Trump having a big advantage.
The second biggest poll- the PPP poll- shows Bush tied for last place with Ben Carson. If that happens, and Kasich is ten full points behind Rubio, a strong case could be made that Rubio would have the "sane lane" all but wrapped up, and barring a stellar showing by either Bush or Kasich (more likely Kasich) on Super Tuesday should, if Bush and Kasich think of the party and the country first, move on from there into what figures to be a fairly even three-way race between Trump, Cruz and Rubio with the strong possibility of a brokered convention in Cleveland. If that happens, Rubio would be the most likely of the three to emerge as the nominee.
Bloomberg's model is similar to that of SC House while Emerson's and Monmouth's foreshadow Rubio emerging in an even more dominant position than does the PPP poll. The story in South Carolina will be which of the two models prevails- a virtual knockout third or even second place finish for Rubio with Bush and Kasich back in the pack, or a close race between two or even all three of the "sane lane" candidates preventing Rubio from establishing himself as the strongest of the three going into the Nevada Caucuses on February 23 and Super Tuesday on March 1.
If the former there is a good chance that Trump can be stopped. If the latter, it still could happen- but those of us who are fearful of a Trump nomination will have cause to begin to worry.