Greetings from a rookie. I am about to promote my own theory, but I do this with respect and self-consciousness!
I have done some work on the statistical connection between flare activity and the motion of planets. This led to a publication and a preprint about modeling and forecasting the solar cycle, and a recent preprint about the timing of individual flares. By "flares" I'll refer to all Earth-facing M and X class flares.
The last four cycles, for which there are continuous records of flares, show a surprising increase in the count of flares towards the alignments of Jupiter and Saturn, and a decrease towards their quadratures, as seen in this plot (which I wish I knew how to resize):
This observation led to some data analysis and eventually to a model which describes solar activity in terms of flares, based on a coupling between the two giants' motion and an internal solar component, presumably of magnetic origin. A description of the model and its development can be found at my webpage here: http://www.chapette.net/solar.html
The peer-reviewed article itself is here: https://arxiv.org/abs/1702.00641
Let me paste the model's reconstruction of the last four cycles and the coming years. This reconstruction used as input only the data of cycle 21, the dates of planetary alignments, and the dates of the start of each cycle (you can find the details in the links above):
The purple histogram is the model and the black is the observations. The reconstruction (again, with minimal input) can be called rather good, but let's now go to the fun part: predictions. The first prediction (uploaded in early 2017) had been the increase in activity in late 2017 - which happened.
The second prediction is planned to make or break the model within the next decade:
Cycle 25 will be weak, spread out, and will consist of two distinct time ranges of activity. To read the details about this forecast, please go either to: https://arxiv.org/abs/1907.06545
or again to my webpage: http://www.chapette.net/solar.html
This forecast was uploaded last summer and predicted that the first flare of cycle 25 will most probably appear in "the weeks following late March 2020". This actually happened on 29th May. If it's not a coincidence, it is encouraging by space weather standards.
For completeness, let me add that knowing the actual date of the first flare led to an update of the prediction, without any significant changes, which is being written up.
And, about the relation to individual flares and possible contribution to forecasting them. More statistical analysis showed strong indications for non-randomness between the timing of flares and the relative motion of the five innermost planets.
For example, this is the distribution of flares as a function of the absolute heliocentric longitude between Earth and Venus:
For more quantitative tests, and also for more indications, please have a look at the very new preprint: https://arxiv.org/abs/2006.10694
A preliminary application of machine learning was also used to see if these relations can be useful in something, even if we don't know the underlying physics yet... it seems that they can. There is an interesting correct classification of 62% of the days in the last four cycles as having flares or not, without using any solar observable.
Finally, what about the next couple of weeks?
The six innermost planets are almost aligning. A quick look at alignments of 5 out of the 6 innermost planets (in the linked paper) is inconclusive. The cycle is at its start so we can't expect particularly strong activity; but if the first two weeks of July turn out unusual, it might have something to do with the above.
I'd look forward to any comments. I thought it better not to write here full descriptions of the related work, but leave this for the discussion if needed.