Here's another take on the effect of RFK dropping out.
The hot take is that with Biden in the race, the polling was 38.6% Biden, 42.1% Trump, and 9.2% Kennedy. (The math says 10.1% undecided or other)
Currently, with Harris, the polls show 46.0% Harris, 43.3% Trump, and 5.2% Kennedy. (Again math says 5.5% undecided or other)
Without Kennedy, it's 49.2% Harris, and 47.6% Trump. (leaving 3.2% undecided)

So switching from Biden to Harris pulled 4% away from Kennedy and another 4+% out of undecided, with over 7% going to Harris and 1.1% going to Trump. Clearly, this change was terrible news for Trump, and since we're in a political forum, I'll add that it's great news for Democrats. Clearly the right move.

Getting Kennedy out of the race cuts into Harris' lead, but not by much, and still leaves things within the margin of error, I suspect. The data are a composite of multiple polls, so there's no stated margin of error. But polls typically have a MOE something in the 3-4% range, with some big national (and therefore sort of useless) polls getting down to maybe 2.5%.

What struck me the most is the drop in the combination of Kennedy and undecideds going from Biden to Harris. My guess is that a lot of those folks weren't all that thrilled with Biden, but simply couldn't vote for Trump. So they split between Kennedy and undecided. Swapping in Harris gave those folks an option they could back. Those who insisted on voting could do their protest vote for Kennedy, and the rest might have just stayed home.

Removing Kennedy from the race is much harder to discern. All of the movements - including undecideds - are likely within the margin of error. So they don't tell us much.

3 minute video from MSNBC
YouTube: Steve Kornacki breaks down how RFK’s potential exit could impact the race (MS NOW)

--Peter