By now, this is 4 year old thread, but the assertion that an absolute pressure transmitter can
 - be configured for negative LRV, -50 mbar
 - would measure without excessive error
should be explained.

Which model is under discussion, absolute or gauge pressure?
The current catalog (2012) does not list a 7MF-4332 model, but it does list a 7MF-4333 model which
 - is an absolute pressure transmitter
 - has an identical range, 8.3mbar a to 250mbar a,
so I'll assume that the model under discussion is an absolute pressure model.

Further to the point, the original poster's (OP) statement, "this is below normal atmospheric pressure" demonstrates that the range 8.3 to 250 mbar is understood to be absolute, and should be stated as 8.3 to 250 mbar a; a for absolute.

The original intention appears to be to get a 50mbar below atmosphere measurement.   The only way to get a reliable reading below atmosphere is to have a reference to atmosphere.  To do so, a gauge pressure transmitter is needed, because a gauge pressure transmitter is referenced to atmosphere.  An absolute pressure transmitter does NOT have a reference to atmosphere; the reference for an absolute pressure transmitter is absolute zero.  A measurement of -50 mbar a is impossible because it is below absolute zero.   A measurement of -50 bar g is possible with an appropriate model pressure transmitter.

As proposed, an absolute pressure transmitter with a sufficient range could be ranged so that it measures pressure above and below an average atmospheric pressure, but
 - the transmitter model in question is incapable of measuring the proposed range; its URV (upper range value) maxes out at 250mbar absolute, which is approximately 760mbar below the 'standard' 1013.23 mbar a atmospheric pressure.
- such a measurement (with a transmitter capable of that range) would have a constant error due to elevation above sea level and a continuous, dynamic error that changes with the barometric pressure and the weather. 

For the relatively low span of 300mbar that was mentioned, such an error would constitute a considerable percentage of the reading, on the order of 10's of percent. 

The historical barometric pressure extremes for Chicago, IL are 971mbar to 1049 mbar, a span of 78mbar.   Extremes are just that, extreme, so halving the span of extremes from 78 mbar to 39 mbar would still produce, over time, an error of 39/300 or 13% of span.   Not exactly an accurate measurement because its reference point, absolute zero, is incorrect, when it should atmospheric pressure (which is what a gauge pressure transmitter does by definition).