Hi,
I like the idea of the new -collect- functions, but in my first application it take >5 minutes to export a single table. During this time the computer fan runs at a very high rate. In Task Manager, Stata is running >30% of CPU and 200MB of memory even though the table is ultimately something like 50 KB. (KB, not MB).
In Windows Explorer, the new file keeps appearing to grow in size up to ~30 KB, then flip to 0 KB in size, then grow again, etc.
Here is a toy example that replicates the behavior. Admittedly it is a moderately sized, sparse table. But this is the nature of the data. It strikes me as a little odd that the table itself is pretty quick but the -collect export- takes forever, and the behavior of the file size when you watch the directory also seems odd.

Code:
clear all
set obs 2500


gen Ab = runiformint(1,35)
gen Resistant = runiformint(0,1)
gen Sp = runiformint(1,20)

table Sp Ab , statistic(count Resistant) statistic(mean Resistant) nototal replace
collect export  "test.xlsx", replace

Thanks,
Scott