In the first place, you should be creating summary data for the summary index. Re-indexing a shorter version of all or even a filtered subset is extremely redundant, unnecessary and not the purpose of a summary index.
All that said, | table * includes many metadata fields that should not be included in a summary index including all the "_" fields. Given that you have said the number of fields is quite large, it is hard to imagine a case where such belongs in a summary index. But if you must, you may want to try adding a | fields - _* to remove the underscore fields and a few others like sourcetype, host, source, index, etc.
Also, ensure that your search head is forwarding to the IDX tier so that you take advantage of the parallelization. Also, make sure that you are using collect instead of tscollect .
... View more