Polls are easily massaged to give a pre-determined outcome and do no always directly correlate with the reality of politics. We've been shown poll numbers like this for years now, and at least at the federal level very little has come to pass.
A poll can be easily manipulated, as 44 AMP pointed out, depending on who you poll and how you ask the questions. For example, in recent years many polls are infamous for drastically over-sampling Democrats and under sampling Republicans. Historically, the split of people who vote Democrat to Republican to Independent/Swing Voter are about 40% Democrat, 40% Republican, and 20% Independent/Swing, but in many of these polls Democrats will be sampled in the mid-to-high 40's while Republicans are sampled in the high 20 to low 30's, that's not an accurate representation of median political ideology nationwide. Also, the demographic and area where polls are conducted can have a large outcome on the result of said poll. If you ask a question of primarily young, minority voters in California, you're likely to get very different answers than if you ask primarily elderly, white voters in Wyoming.
Likewise, how you ask the question will have a lot to do with the responses you get. For example, in poll after poll the majority of respondents are in favor of "background checks" but you often find that that favor drops off sharply when the specifics of what a new "background check" law might entail. Once people find out that they'll be criminals if the loan a family member a gun or try out someone else's gun at the range without first going through a "background check" the new law suddenly seems like less of a good idea.
Finally, the results of a poll don't necessarily correlate to what a given politician will do. The only federal office that is subject to a nationwide election is that of President. Congressmen and Senators are subject only to elections in their districts and states. I can pretty much guarantee you that Congressman Smith or Senator Jones is a whole lot more interested in the opinions of the constituents in his/her particular district than the result of some nationwide poll. A more accurate measure would be to compile polling data by individual districts, but that's a lot more involved and doesn't make for a good headline or soundbite so I don't expect the media to do it.
You also have to take voter enthusiasm into account. Just because someone answers that they support or oppose something in a poll doesn't mean that it will be the foremost issue on their mind when they cast their vote, assuming they vote at all. Historically, people who oppose gun control, on the whole, feel much more strongly about it and are more likely to to vote for or against a certain candidate because of it than people who support gun control. Most people are not single-issue voters and among those who are, gun control isn't usually the single issue. I suspect that the party currently in power wants to talk about gun control right now (or the myriad of other #TheCurrentThing issues we've been bombarded with of late) because they very much don't want to talk about product shortages, inflation, and record high gas prices as, on their handling of those issues, they aren't looked upon very favorably by most voters.
A poll can be easily manipulated, as 44 AMP pointed out, depending on who you poll and how you ask the questions. For example, in recent years many polls are infamous for drastically over-sampling Democrats and under sampling Republicans. Historically, the split of people who vote Democrat to Republican to Independent/Swing Voter are about 40% Democrat, 40% Republican, and 20% Independent/Swing, but in many of these polls Democrats will be sampled in the mid-to-high 40's while Republicans are sampled in the high 20 to low 30's, that's not an accurate representation of median political ideology nationwide. Also, the demographic and area where polls are conducted can have a large outcome on the result of said poll. If you ask a question of primarily young, minority voters in California, you're likely to get very different answers than if you ask primarily elderly, white voters in Wyoming.
Likewise, how you ask the question will have a lot to do with the responses you get. For example, in poll after poll the majority of respondents are in favor of "background checks" but you often find that that favor drops off sharply when the specifics of what a new "background check" law might entail. Once people find out that they'll be criminals if the loan a family member a gun or try out someone else's gun at the range without first going through a "background check" the new law suddenly seems like less of a good idea.
Finally, the results of a poll don't necessarily correlate to what a given politician will do. The only federal office that is subject to a nationwide election is that of President. Congressmen and Senators are subject only to elections in their districts and states. I can pretty much guarantee you that Congressman Smith or Senator Jones is a whole lot more interested in the opinions of the constituents in his/her particular district than the result of some nationwide poll. A more accurate measure would be to compile polling data by individual districts, but that's a lot more involved and doesn't make for a good headline or soundbite so I don't expect the media to do it.
You also have to take voter enthusiasm into account. Just because someone answers that they support or oppose something in a poll doesn't mean that it will be the foremost issue on their mind when they cast their vote, assuming they vote at all. Historically, people who oppose gun control, on the whole, feel much more strongly about it and are more likely to to vote for or against a certain candidate because of it than people who support gun control. Most people are not single-issue voters and among those who are, gun control isn't usually the single issue. I suspect that the party currently in power wants to talk about gun control right now (or the myriad of other #TheCurrentThing issues we've been bombarded with of late) because they very much don't want to talk about product shortages, inflation, and record high gas prices as, on their handling of those issues, they aren't looked upon very favorably by most voters.