Note that these Ipsos numbers are from their tracking poll, not the post-debate poll they ran yesterday showing the race going from Obama 48-39 pre-debate to Obama 48-43 post-debate. Those were one-day samples, not the rolling sample here.
Remember, the Ipsos rolling sample is five days long, Rasmussen's is three days, and Gallup is seven days. So the post-debate sample is just
1/5th 2/5th (see update), 1/3rd and 1/7th of those numbers, respectively. But had Obama tanked the day after the debate, there would be some slippage in the tracking poll toplines, not improvement for Obama. So if Romney saw any bump in his Thursday polling, it was offset by even better Obama polling before the debate.
One thing is true—oftentimes it takes several days to a week for a big campaign event to be reflected in the polls, as news junkies disseminate the information at water coolers and backyard picnics across the country. That is particularly the case in tracking polls with rolling samples, so we need to keep watching over the weekend to see if any latent damage starts showing up.
However, Republicans don't seem to be winning the post-debate spin, not when the "moment" that came out of the debate wasn't anything Obama said, but the fact that Mitt Romney wants to fire Big Bird.
Indeed, it wasn't the Romney campaign who is turning debate moments into television attack ads and internet memes.
Update: A new Ipsos/Reuters poll has been released, and I updated the chart above accordingly (the previous one had Obama +1). So Obama has held stead and Romney has gained +3. FYI, the new Ipsos numbers include data collected today, so it now has two days of post-debate responses.