The alternative approach taken in a project described in [#!unixl3!#] is to provide an emulation of LibMach which provides a veneer over the Mach kernel. The conclusion of that report is that providing a Mach emulation on top of another microkernel is unnecessarily complicated and it is probable that altering the overlying operating system to work with L4 directly would be significantly faster.
This does not necessarily mean that a common microkernel abstraction layer such as that which LibMOM attempts to provide is going to be inefficient. Much of the overhead associated with the LibMach approach was consumed in emulating the exact semantics of Mach. This would not apply to LibMOM since it implements very simple primitives.