As a software engineer, I can say that if it is a relatively obvious mistake (like the recent 501/508 bug that caused the DST fix to be apply after every reset, instead of once a year), you can change the one line, recompile it, do some tests, and get it out in a day or two (which they did for that 501/508 bug).
BUT, if it is a more sophisticated problem, like this two tuner problem, then it is not immediately obvious what needs to be done, so certainly the CSRs are not going to know how long it will take.
E* (and most other software development departments) don't like to revert to the previous version, partially because it looks bad from a marketing viewpoint, but more because the time spent reverting could be more profitably spent fixing the bug, which will have to be done anyway.
E* in particular has unique problems with reverting to older versions, because in addition to cute interface features, the new software often has essential support features, such as adding a new satellite position (like 105, 121, 118.75, or 129) that is absolutely needed just to provide channels.
Generally, E* has a good track record of eventually fixing all the problems, and they are more accessible than most engineering departments - word about problems really does reliably reach them.
Their resources are not infinite
, so some patience is necessary. If you have two receivers, it often helps to have two different models.
