Those who say the issue makes them less apt to back Clinton overwhelmingly back Trump in the first place. But what's key in a close contest is whether the disclosure gives Trump supporters further motivation to turn out for him and whether it demotivates reluctant Clinton backers.
In other results of the survey, produced for ABC by Langer Research Associates:
• Regardless of the closeness of current preferences and plans to participate, 60 percent of likely voters still expect Clinton to win. This too may be a risk to her, possibly leading some Clinton supporters not to bother voting.
• Clinton continues to top Trump in two key attributes. She is seen as more qualified to serve as president by a substantial 54 percent of likely voters, versus 36 percent for Trump, a remarkable result, given the closeness of the race overall. She's also seen to have a stronger moral character, albeit by a closer margin, 46 to 38 percent.
In earlier ABC News tracking poll results, a broad 60 percent of likely voters disapproved of how Clinton has handled the email issue, including 3 in 10 of her supporters and 93 percent of Trump's. Nearly all of them disapproved strongly, and that was before Comey's disclosure Friday.
The possibility of a pullback in motivation of Clinton supporters or further resurgence among Trump's may cause concern in the Clinton camp — especially because this dynamic already was underway. Intention to vote has grown in Trump support groups in the past week as the intensity of criticism of him has ebbed, including allegations of sexual misconduct, disapproval of his position on the election's legitimacy and his poorly rated final debate performance.
As noted, Clinton leads Trump on qualifications and moral character but not significantly on empathy. At this point in the 2012 campaign, Obama had just opened a 6 point advantage in better understanding people's economic problems, and it was 8 points at the close.
Clinton, as it happens, had an 8-point lead over Trump on empathy in early September and 10 points in early August. She has lost 13 points on the measure among liberals, a core Clinton support group, since Sept. 8, and Trump has gained 10 points among independents.
Clinton led Trump by 12 points in the first four days of the ABC News tracking poll, with consistent results across those nights; that polling followed two particularly difficult weeks for Trump, beginning with the Oct. 7 disclosure of an 11-year-old videotape in which he crudely described his sexual advances toward women.
Results have been equally consistent and far closer across the subsequent five nights. The tracking poll customarily uses a four-night average of results. The average across all nine nights, for comparison purposes, is 48 to 42 percent.
Langer Research diagnoses results across a range of likely voter models. Fifteen such models currently are in use, projecting turnout estimates ranging from 43 to 70 percent of the voting age population. All show essentially the same results using the last four nights of data.
The tightening of the contest the past week, while unusual, is not unheard of. Among many gyrations in ABC News and ABC/Post pre-election polls in the 1992 election, Bill Clinton went from an 11 point lead to a scant 3 point edge against George H.W. Bush toward the end of their race before regaining the advantage in the closing days. (There was also a steep shift in 1996.)
Notably, among groups, a slight Hillary Clinton advantage among independents in early tracking is now a 19 point Trump lead. Part of the reason: a decline in the share of Democratic leaners within the ranks of independents who are likely to vote.
The latest results show a continued but different gender gap, from Clinton up 19 points among women and up 2 among men in early tracking to Clinton up 8 points among women and Trump up 9 among men now.
There's a much closer race, meanwhile, among young voters, again because of signs that fewer Democrats and Democratic leaners in their ranks are joining the likely voter pool. Bernie Sanders was especially popular among young adults, and they've been far less strongly enthusiastic Clinton supporters than their elder counterparts.
Trump has a 15 point advantage over Clinton among whites, while Clinton has 68 percent support in the latest results among nonwhites, versus Obama's 80 percent among nonwhites in the 2012 exit poll. She's backed by 88 percent of blacks (average for a Democrat) and 60 percent of Hispanics (a bit less than average; 10 percent peel off to Johnson and Stein). A notable difference is nonwhites who are not black or Hispanic. Obama won them by 66 to 31 percent; it's a closer 45 percent for Clinton and 40 percent for Trump in the tracking poll.
White Catholics, a swing voter group in elections from 1976 to 2004, are now with Trump, 57 to 36 percent, reopening his advantage in this group to near the margin he last saw in late September.
Trump counters Clinton's 21 point lead among college-educated likely voters with a 12 point advantage among those who lack a college degree. (The latter are more numerous.) Clinton leads by 23 points among college-educated white women, a crucial bulwark for her. The two are even among college-educated white men, while Trump leads by 38 and 27 points, respectively, among non-college-educated white men and non-college-educated white women.
Non-college-educated white men have been a steady Trump group. Non-college-educated white women have been for him generally but by less consistent margins. Where they end up may well prove critical in this still unpredictable election.
This ABC News/Washington Post poll was conducted by landline and cellular telephone Oct. 25 to 28, 2016, in English and Spanish, among a random national sample of 1,160 likely voters. Results have a margin of sampling error of 3 points, including the design effect. Partisan divisions are 37-28-30 percent, Democrats-Republicans-independents.
Question 6 was asked Oct. 24 to 27 among 1,268 likely voters; those results have a 3 point error margin. Question 7 was asked Oct. 28 to 29 among 480 likely voters; those results have a 5 point error margin. Question 8 was asked Oct. 26 to 28 among 415 Republicans and Republican-leaning independents who are likely to vote; those results have a 5.5 point error margin.
The survey was produced for ABC News by Langer Research Associates of New York City, with sampling, data collection and tabulation by Abt-SRBI of New York City. See details on the survey's methodology here.