OK, scratch the above Nero example, at least with a Lite On SlimtypeDVDRW SSW-8015S. I burned a disc with Nero (no 2 second gaps) with this lineage:

source: DSBD->DAT@48K
transfer: DAP1-> echo MIA-> SF7.0-> CD-Wave-> FLAC

This fileset had no SBEs or other errors.

My drive has a read offset correction of +594 and a write offset correction of -6. So I used a combined read/write offset correction of +588 and extracted the last track of disc 1 and EAC reported missing sample errors in the log file. Then I tried extracting a track from the middle of the disc, combined read/write offset +588, got missing sample errors in the log file. I tried a commercial CD, combined read/write offset +588, got missing sample errors in the log file. Then I tried the commercial CD again, read offset +594 and got no missing sample errors. Then I tried the burned CD again with combined read/write offset +588 with "overread" enabled and got missing sample errors. So using a read offset of anything other than the correct +594 in my drive/ASPI interface/laptop results in missing sample errors.

I remember extracting with the combined read/write offset working fine on an old HL-DT-ST CD-RW drive I had on a different computer though.

Last thing, when you extract a Nero burned CD with the read offset corrected, the missing sample errors won't show up in the log file of EAC, because they aren't read errors. The missing samples will show up when you are using the compare WAVs feature of EAC in Tools > Compare WAVs. I'm not sure what the difference is between these two types of 'missing sample' errors.