TACA incorrectly assumes all samples are miseq if mosfs is down

[sylvinite]

TACA incorrectly assumes all samples are miseq if mosfs is down.
This can mean things like TACA creating incorrect database entries based on the assumption that it has found a miseq run.
This could be resolved by either having a more stringent miseq run pattern (whilst still accomodating to lab deviations); or by simply not having TACA run at all if /srv/mfs isn't found.

Regardless it probably has to be discussed before something is implemented.