This implies that pack production is still a bottleneck
Harder to say that exactly... they can produce a maximum of 5,000/week without the Grohmann line, they said that during the call (or in the ER). The Grohmann line is more automated and simply reduces the number of staff required to make the 5,000/week - thus, it is less costly. But it won't go faster than 5,000/week yet.
I wonder if the "Grohmann line," as it is designed & built today, will
never be able to produce more than 5,000/week. I suspect that their push from 5,000 to 10,000/week will be accomplished by installing another line. Don't forget that GF1 is only one-third built, and they have already described it as having multiple identical "production lines." The pilot line could be considered complete when it is reliably producing 5,000/week. After that they put in another set of everything to get it up to 10,000/week. Does this make sense?
I don't think they have enough room to stockpile excess produced parts from areas/processes of the factory that are going faster than the next area/process. As they improve, the following area/process would then have to not only catch up with its input areas, but surpass them, in order to absorb the stockpiled stuff (or the input processes slow down). And finally, the slowest area/process kinda dictates the speed of overall production, even if the following areas/processes
could go faster - they simply run out of materials to work with, because they're waiting for the slower input departments. This is exactly what Elon has said on many occasions - production goes at the speed of the slowest or least-lucky department.