Those options are:
1. Give out source to the library, and either source or object code to the work that uses the library (so that the work can be relinked)
2. Use shared linking and build the work into a .so that would work if the user replaced it (There is a point of contention in LGPL 2.1 as to whether you need to make it possible for the user to replace the .so. It's clear what the spirit is, but the actual wording does not require it)
3. Include a written offer for the stuff in #1.
4. Offer source from #1 through the same place that offers binaries.
5. Verify that the user already has the source, or that you already sent it to the user.
If the app store requires static linking, that rules out #2.
""" For an executable, the required form of the "work that uses the Library" must include any data and utility programs needed for reproducing the executable from it. However, as a special exception, the materials to be distributed need not include anything that is normally distributed (in either source or binary form) with the major components (compiler, kernel, and so on) of the operating system on which the executable runs, unless that component itself accompanies the executable.
It may happen that this requirement contradicts the license restrictions of other proprietary libraries that do not normally accompany the operating system. Such a contradiction means you cannot use both them and the Library together in an executable that you distribute. """
I believe this makes #1 insufficient, (and by reference, #3, #4, #5) - as that requires apple to also distribute the material needed to sign the executable for your specific phone (as an end user) if they give you the executable signed for your specific phone.
Do you have reason to believe otherwise?
You may want to get into rehashing that discussion, but i've already had that discussion more than once during GPLv3 and LGPLv3 drafting, and it was enough fun there :)
I also am firmly in one of those camps of open source lawyers, and so it wouldn't be appropriate for me to try to reproduce the arguments of the other side, but, at the same time, nothing is ever as black and white as most folks like to think, and I would be stupid to pretend they didn't have a reasonable argument.
In any case, this is why GPLv3 and LGPLv3 is explicit on this point.
I can't answer your question, but just point out that huge chunks of iOS itself are LGPL already, such as webkit, and the LGPL version of ffmpeg is already widely used.
There has been some speculation in the past that because iOS mandates static linking that this is a problem for using LGPL libraries, but I fall into the 'linking creates a compilation, not a derived work' camp and if the compilation interpretation is correct then static linking isn't a problem.
But every time I dig into LGPL interpretation and linking issues, eventually my head starts hurting and I give up.
2. The rest of what you said doesn't make much sense to me. I'm not clear on what distinction you are trying to make.
There are those that believe that the question of linking is mostly irrelevant to whether something is a derivative work. Lawrence Rosen is the canonical example of folks in this camp. But even in that camp, static linking may be a problem, it just depends on what is being done.
So it's not really true that "static linking isn't a problem".
Specifically, it cares that the end user can "swap out" the LGPLd code for their modified version, which is not available in the iOS AppStore regardless of compilation vs. derived work interpretation (a distinction without difference if I understand it correctly).
How does the LGPL relicense solve the Apple AppStore problem everyone is talking about? Apple may approve LGPLd code, but as far as I can tell, they still do not comply with it under their current distribution system.
edit: wrote iOS instead of LGPL in the first line, thanks simonh