Here's some quick math. With the updated Ontario results McMurray is 26,231 votes behind Jacobs. Considering McMurray won absentees in the most Republican county in the district it's pretty safe to say he can only gain on Jacobs from here. The question is by how much.
In the three counties reporting absentee results as of now, Jacobs' percentage of the absentee vote is 35.3, 34, and 33.8 points lower than his percentage of the election day vote. Let's average that out and say Jacobs will do 34.4 points worse in the absentee ballots than election day.
Applying that to the number of absentee ballots uncounted in the 5 outstanding counties:
Erie (~41000): Jacobs won 69.7% of the ED vote so he'll drop to 35.4% of the absentees - with a third party vote of 1.6% McMurray will net 11316 votes
Niagara (~12000): Jacobs won 66.6% of the ED vote so he'll drop to 32.2% of the absentees - with a third party vote of 1.6% McMurray will net 4080 votes
That brings McMurray within 10835 votes of Jacobs. I haven't been able to find absentee ballot statistics for the three remaining counties (Orleans, Genesee, and Monroe), but together in 2018 they cast about 66% of the total Ontario + Wyoming + Livingston vote so let's say they cast 66% of the total Ontario + Wyoming + Livingston absentees which would be 9797 ballots in total - even if McMurray won them with 100% of the vote it wouldn't be enough.
With all those assumptions and calculations McMurray would probably between 6000-8000 votes behind Jacobs. What McMurray needs are incredible showings in the Erie and Niagara absentees - better than the ones he's put up in the three counties that have already reported. If McMurray were to somehow get 70% of the Erie absentee he would net an extra 5740 votes - repeat that in Niagara and that's probably enough for a narrow win.
I had 7400-5900 on my spreadsheet for Ontario in terms of a fairly generic result lying under the assumption that Jacobs vote total in the absentees are equal to the GOP ballots returned and then add the remaining number of ballots to McMurray's total. For example, Ontario County reported 1922 GOP ballots and 6734 total ballots, so I added 1922 to Collins' total and about 4800 to McMurray's total, but this didn't account for the extra influx of ballots received after election day. This result lines up well with a similar trend in Livingston and Wyoming (although if I recall correctly there were fewer GOP votes counted than GOP ballots returned in absentee, probably by about 3% or so).
Using this method to predict the remaining counties demonstrates how difficult it is for McMurray to win at this rate, since as I previously mentioned, the election day vote deficit is probably too large for McMurray to feasibly be able to take the lead in Niagara or build up a margin in Erie County.
Orleans County: 2312 ballots returned as of June 23, 1110 GOP ballots returned, add 1110 to Jacobs total, 1400 to McMurray's total (account for more ballots coming in), get margin of about 29% which is considerably worse than 2018 where McMurray lost by 22.3%)
Genesee County: 3554 ballots returned as of June 23, unknown amount of GOP ballots returned. Will just assume 3700 ballots swinging 65%, final margin around 22%, again considerably worse than losing by 13.3% in 2018
Monroe County: no clue, I'll be generous and just say McMurray wins 3300-3100 overall, slightly better than 2018 marginwise
Niagara County: 3784 GOP ballots, add that to Jacobs total, give 7748 to McMurray (probably might be a couple hundred more countywide in reality still since their numbers are as of June 26). final result 14307-12571 in favor of Jacobs, 6% margin, worse than 1% McMurray loss in 2018
Erie County: 12413 GOP ballots, add to Jacobs total, 28710 to McMurray, 38586-36294 in favor of McMurray, slightly worse than winning by 5% in 2018 (3% using this projection).
Overall this results in a 3.3% Jacobs win, and it might even be narrower since Ontario and Livingston Counties have both produced fewer Jacobs votes than GOP ballots cast. The issue that prevents McMurray from having any reasonable chance is that Jacobs would have to receive around 10-15% fewer votes than GOP ballots cast for McMurray to have a shot at winning, and there is no evidence to suggest that may be a possibility from the counties currently reporting.
source on Orleans/Genesee absentees (have to scroll down a bit to find it)