What about Gradient Lag?

Can you use a little bit of common sense? I'm obviously talking about the 1040, which atm, has more important issues to be addressed. Don't care how old this 'bug' was around when your device crashes, freezes, or doesn't save activities.

Also, it does depend what you call "lag", according to Garmin if it's just ~10s then it's normal so.

The problem is that Garmin call 10s lag as "normal", like it's interesting to always show what the gradient was 10 seconds ago. When I'm using my common sense my conclusion is that that is ridiculous. I also think that the lag on the 1040 have the same cause as the lag on 820.

If Garmin says they are working on it that would be fine, but what most people in this thread is reacting to is that Garmin says "it's within expectations". At least I am worried of what their "expectations" is for other functions, and for new or changed functions in the future like the gradient, that was fine on Edge 800.

Actually, 15 s seems perfectly reasonable when you consider how Garmin is (currently, at least) calculating gradient.  They measure the distance traveled using either GPS or a speed sensor and measure change in height using the barometer.  They have to measure for a long enough time to get an accurate measure of the gradient given the noise in the two measurements, bumps in the road, etc..  Suppose you want to measure to 0.5% grade accuracy in 5 seconds while going 10mph on a 10% grade.  In 5 seconds you would travel 73 feet and rise 7 feet.  To get 0.5% grade accuracy you'd need the rise measurement noise to be less than 4.4 INCHES of elevation change.  Which is a tiny value for a barometer.  If you say 15s instead then you are talking about 1 foot of elevation noise.  For a barometer with a full scale range of, say, 15000 feet that corresponds to a noise level of 0.007% of full scale which is great for any measurement system.  Even if you change from a barometer to a gyro or something like that you still have to contend with measurement accuracy of that system.  Garmin could measure for a shorter period of time but that would result in worse accuracy.  Most people wouldn't like their grade to read 10% with random +/-2% noise.  This is the same method that everyone else uses.  You trade measurement time for accuracy of measurement.  Garmin has decided that most of their rider are happy with 5 second response time to give reasonable accuracy.  I actually don't care about gradient at all and don't know anybody personally that does.  So, my guess is that the status quo is perfectly fine for the vast majority of Garmin's users.

+/-2%  on 10% - 20% would be perfectly acceptable and would capture most hills.  I’d personally rather it was quicker.  What frustrates most of us on here is Garmin not accepting the problem when in their previous generation devices it was fine. 

At least give us the choice of algorithm in a setup parameter rather than telling us they don’t care. < 0.5% in terms of accuracy whilst riding, for me, would be fine as long as it’s fast. 

For you, maybe it’s not important. For others, not unreasonably, it is.

After acknowledging that my 1040 shows actual gradient 10 seconds late, I always count 10 seconds from the point I want to know the gradient, or do the other way around, I watch the gradient, then go flashback, so I get an idea of what was the gradient 10 seconds ago. You can use a tree, bench, or any other obstacles to use as mark.

10 seconds is tolerated if you use your common sense as soon as the gradient shown up is accurate'ish. Same thing goes to denivelation. I  think we're just making fuss out of it, however some claims having a 30 seconds delay, which is definitely not normal.

How responsive and accurate was it on Edge 800 anyways?

We can argue all day that 15 second delay is acceptable or not. However the main issue is that this wasn't a problem on the older 1030 and it has been an issue since the 1030+. If you really want to get this fixed or at least more attention from garmin is to post some videos on YouTube or any other social media outlet with a 1030 and a 1030+, 1040, 1040 Solar and show the grade lag that way. Posting here on Garmin forums is landing one deaf ears. Social media will get this going. You can currently search Youtube for anything related to this issue and there isn't anything. I would do a video however I sold my 1030 before I could get around to making a video. MAKE VIDEOS PEOPLE!!!! POST THEM ON OTHER SOCIAL MEDIA PLATFORMS!!!!!

I disagree:

a) I would trade speed for accuracy. Having said this, there are many (cheap) inclinometers that measure it instantaneously within reasonable accuracy (just goggle it) that would be better off than the one provided in the Edge 1030+/1040. It is a matter of using that "older" technology instead.

b) As for noise, +/-2% would be much better than the actual readings. I understand it is not the same but for example having a reading of +17% when I am over the top and going down at -2% is much worse that a reading between 0% and -4% at the same spot and moment, which is what happens now.

c) In regards to people that care, maybe your acquaintances don't care about this parameter, but if you see the forum and social media, you will see hat there are many persons that do. Most coming from older devices that expect improvement and not getting worse readings than in the past. Also, it is important to note that some people care about the recorded values and some about the instant values. The recorded values can always be corrected with the maps info. The instant values are good for reacting and guiding yourself while riding.

Then use Climbpro if you want to pace yourself depending on gradient, these newly added colors can surely be more useful than gradient itself

There are IQ widgets that do exactly that. I tested one of them last week, it was faster and less accurate. In the end I preferred to see a solid reliable value rather thank a faster unreliable one. DC Rainmaker did some comparison between different suppliers and he reported that all of them have different tradeoffs between speed and accuracy, nobody does it right. 

Regarding C) I come from the 1000 that used to be undeniably faster and in my opinion even accurate. While worse on grade reporting, the 1040 has so many most interesting and useful features that I don't really care about this. I'm curious about using grade reporting for reacting and guiding, while I like to see the actual gradient I'm always using my senses to decide the proper gear and effort... grade always come later and you need to prepare in advance for a climb. Maybe it's just me.