I am getting problems too on my S30 which are much, much worse after this latest update - very similar to yours. Sometimes frames in live view will look like yours with extreme trails (or the first frame or two shown in "enhancing"/live stacking mode look like this).
They sometimes even out, or alternatively just keep getting discarded entirely and I have to stop and move to another target.
I tested a lot last night with the new firmware in a rare crystal-clear sky in the UK - Bortle 4/5.
I use EQ mode exclusively with 60-sec exposures (always have done without issue since I got the scope in December) and have a solid polar alignment (I never trust the "refresh deviation" and instead go right out to the advanced menu then back in and let it re-take the images around the zenith and re-adjust if necessary - repeated 2 or three times until the deviation is within 0.1-0.2).
On some targets I have got slightly different results: everything looks perfect for the first 5 minutes with no discarded frames, but then it starts rejecting frames like crazy and rarely accepts any more. Because there's no way to see a "live" view when in live-stacking mode nor a proper reason given by the app (just "stacking failed"), it's hard to say what the problem is but I expect yet another software bug introduced in this latest update. I have submitted a bug report from within the app.
It is considerably worse for objects in the western sky (I am at 52 degrees north) - super high rejection rates for targets there such as the Tadpoles last night, certainly more rejected than accepted over time.
I have been also getting "target offset detected. Re-centering" errors again on objects in the western sky usually when I have slightly changed the target frame in the app to have the object I'm interested in more centred (the default object definition locations in the Sky Atlas are often not dead-centre to the S30's FOV, and can be subject to the scopes inherent coma if too close to the edge of the FOV unless you do this). Been doing this the whole time before the latest update without problems.
Strangely though, I pointed it to the Leo triplet (which was almost due south last night) and got 30x 60-sec exposures continuously with zero dropped frames so it must be something to do with when the scope is in a specific orientation.
I notice the scope has always been able to "wobble" minutely around the centre tripod axis left/right even when fully tightened up which I hear from others is totally normal and is assumed gear backlash, and I presume this is now not being corrected for (or over-corrected for) when imaging objects in certain parts of the sky. I think once the scope hits a certain angle it may slump down slightly.
This thread (https://bbs.zwoastro.com/d/25831-seestar-s30-tracking-issues-above-55/8) mentions this phenomenon too and that adding weight to one side of the scope to force it to have something to slightly "push against" to stop it slumping down resolves it.
I had this same issue with western targets it in the last firmware but to a much lesser degree than now - last night it was very very hard to get a good set of data without way more rejected frames than accepted ones.
Here's another thread: https://bbs.zwoastro.com/d/26285-s30pro-is-dropping-all-frames-in-eq-mode/18
There are others on the official ZWO Seestar Facebook group complaining of increased dropped frames too since the latest update.
