No doubt, others have already done this, but in case they haven't....
Another how many approach to XPs is the use of mathematical extrapolation of the discovery rate.
Assuming discovery technology only maintains the rate and does not accelerate it - the easy ones, afterall, are limited in supply - then the following XP number is what might be awaiting us...
Code:
Year Discoveries Total
1989 1 1
1990 0 1
1991 0 1
1992 3 4
1993 0 4
1994 1 5
1995 1 6
1996 6 12
1997 1 13
1998 7 20
1999 10 30
2000 19 49
2001 12 61
2002 34 95
2003 26 121
2004 29 150
2005 32 182
2006 28 210
2007 50 260
2008 83 343
2009 79 422
2010 91 513
2011 105 618
2012 132 750
2013 147 897
2014 192 1,088
2015 238 1,327
2016 295 1,622
2017 361 1,983
2018 407 2,390
2019 478 2,868
2020 601 3,469
I used 50 for 2007 as a guess based on 31 so far for the year. I don't have the months of discovery for comparison. Are more announcements made earlier than the year. If, however, the latter part of the year is about the same as the first half, then 62 should could be a better number. Using 62 for 2007, then we have...
Code:
Year Discoveries Total
1989 1 1
1990 0 1
1991 0 1
1992 3 4
1993 0 4
1994 1 5
1995 1 6
1996 6 12
1997 1 13
1998 7 20
1999 10 30
2000 19 49
2001 12 61
2002 34 95
2003 26 121
2004 29 150
2005 32 182
2006 28 210
2007 62 272
2008 90 362
2009 87 449
2010 102 551
2011 120 670
2012 154 824
2013 175 1,000
2014 233 1,233
2015 296 1,529
2016 372 1,901
2017 459 2,360
2018 517 2,878
2019 631 3,509
2020 806 4,315
Interesting, but I don't intend to suggest it is likely, or unlikely, just interesting.
[Edit: I quickly realized I had an error and this is the corrected extrapolation]
[Added: Shoot!, it is wrong, too. It is stricly linear! Dang. I used linear interpolation (all I found on Quatro Pro).
It is now based on the Growth function. Hopefully, this is more mature.
Source is:
http://vo.obspm.fr/exoplanetes/encyclo/catalog-all.php ]