That’s odd. 5 seconds is well beyond anything I’ve seen in tests.
My general rule of thumb for the standard 30x30 trainer test I’ve done for every trainer/smart bike in the last forever years is 2-3 seconds to fully stabilized (no overshooting). 4 seconds would be on the outer edge of that. As you noted, too quick and it’s the brick wall effect (e.g. 1 second). 2 seconds is fine though in most cases. The test is done going from ~150w to 450w, with steady cadence at 90rpm.
As others noted up here, you can see TrainerRoad trigger the interval about 1-second early, depending on the exact trainer model (I account for this 1-second in my test, and time it based on when TR starts). Here’s an example of the Decathlon D100 trainer, and you can see it trigger it 1 second early, and stabilize 3 seconds later.
That said, it’s always been interesting discussing ERG mode between TR & Zwift. Years ago I pointed out to Zwift that ERG mode often felt a bit more wobbly there (rubber-band-ish). They stated back they simply follow the FE-C spec (at the time), and nothing more. TR has largely said the same. Aside from triggering an interval early (or, on their own timetable), I’m not aware of any “special sauce” any of these apps can do with respect to ERG mode on either ANT+ FE-C or BT FTMS. They’re simply sending the wattage set point. Now, again, that doesn’t mean apps will send the set point you think. For example, TR sends theirs a second early. And up above there’s thoughts about Zwift basically faking a ramp in there (thus, sending intermediate set points). But from a protocol standpoint, these apps are all sending the same thing.
That said, in recent days, there does seem to have been some sort of change on the Zwift side, at least in terms of the activation of ERG mode. I’ve watched my wife struggle immensely on every workout over the past week, where it’s not engaging ERG mode at the very start of the workout for some odd reason, unless she hard spikes the power. Same trainer she’s been riding for months without issue.
Likewise, to that extent, a HUGE factor that many people currently overlook when it comes to ERG responsiveness is simply your gear ratio. Especially on stability, and even-moreso now in the era of the Zwift Cog and/or more common 1x setups. One has to really look at those factors, especially on trainers that are either a bit underpowered, or, have different internal flywheel ratios. For example, the Direto XR Cog is rather challenging for me to use on a 1X bike for lower-floor recovery sections. Inversely, the D100 is challenging to use on a 2X bike if you’re not in the big ring (not enough flywheel speed). This isn’t limited to Zwift at all, and is applicable to TR and any other app too.
