Measuring 5'-nucleotidase activity in a cell extract the main problem is to identify the isoenzyme you are measuring, since there are at least 7 different enzymes that can hydrolyse purine monophosphates. Any way if you wont to avoid the kit with MESG, I think that the best way is to measure the conversion of IMP into uric acid  at 293 nm utilizing PNP and xanthine oxidase as ancillary enzymes (both are commercially available). The protocol requires 0.1 M phosphate buffer pH around 7, IMP at your convenience, (around 1 mM) and a few microL of both the ancillary enzymes. You should control your protocol measuring nucleotidase activity as function of the volume of extract added to the assay mixture. There should be a linear relationship at least in a low range, if the range is to low add some more of the ancillary enzymes. The delta epsilon mM between IMP and uric acid at 293 nm is 11.15, this method is sensible but of course, as the kit, does not distinguish between a generic phosphatase and the family of 5'-nucleotidases. Good luck, let me know if you need some more help

Phosphate production can be monitored continuously by fluorescence using this Phosphate Sensor product:

https://www.thermofisher.com/order/catalog/product/PV4407

It's quite expensive, so if you want to use it I would suggest finding a 384-well fluorescence plate reader and some low-volume black assay plates and doing the assay in the smallest possible volume.

Here is another continuous phosphate assay reagent:

https://www.aatbio.com/products/21660

Martin Webb has published several papers on fluorescence-based phosphate sensors based on phosphate-binding proteins with mutations and attached fluorophores. Here is one recent paper:

http://pubs.acs.org/doi/full/10.1021/acs.biochem.5b00449