I didn't fully understand your argument but I already see some problems with it. First of all, the changes in Smith's rating reflect how his skill changed, not rating inflation. Secondly, I don't understand how you came upon 1800 being his rivals' rating. It does seem far too low.
Philidors most estimated rating is 2000!

im not saying its correct but its the closet evidence can point to and yes his closest rival was 1800 by his pattern in rating its probably 1900 but I'm not sure.

1. So who was this 1800-level closest rival and where's the evidence of his 1800-strength?
2. Did you account for rating inflation?

Chessmaster 102 I suggest you read The Ratings of Chess Players Past and Present by Dr Arpad Elo in order to find out at least a fraction of what you don't know about trying to rate chessplayers from a few centuries ago. You might also try checking out Dr Elo's credentials. For starters, he was a Professor of Physics (implying he had a heavy duty math background) and a chess master. I don't believe he tried rating Philidor or any players prior to Morphy, but it's been awhile since I read his book
Of course, everyone's entitled to their opinions, it's just that some opinions are way more authoritative than others
http://www.chessbase.com/newsdetail.asp?newsid=4326
heres my evidence accoding to edochess.ca the oldest chess player being smith (captin smith) had a rationg of 2027 but directly clicked on his rateing shows more of a pattern of going up and down by 7pts but the rating starts from 1859 but philidor played chess till 1795. and according to andrew soltis (his opinoin but more theory) he was 200 pts ahead of his closest rivials. 1859-1795=61x7=427-2027=1600+200=1800+200=2000 I got the additional 200 from the fact that on chess.com he has 20 notable gmaes so i gave 10pts to each(in some masters views this is sort of accurate) So congrats if your a expert 2000-2199 your just as or stronger than philidor even in his prime.