If by 'old' you mean relatively conservative, then yes. But, the same might be said about the Italic languages, with which the Celtic languages share a number of sound laws. If you disregard the fact that they are Satem languages, the Baltic languages might be also considered conservative, however.
Anyways, Celtic shares with Germanic, Greek and Italic to have merged the so-called 'palatovelar' series of PIE (k´, g´ and g´h - which become s, z and ʒ-like sounds in the Satem languages) with the plain velar sounds (k, g, and gh).
k´ > k
g´ > g
g´h > gh
This is known to have occured after the early 3rd millennium BC (common Germanic-Baltoslavic forms which predate the Centum/Satem split and must have arisen in Corded Ware times), but before the mid-2nd millennium BC (attestation of Mycenean Greek, the oldest attested Centum language).
One common sound law is the occurence of *p to *kw before *kw. For example:
PIE 'Penk´we'
Latin 'Quinque'
Old Irish 'Coic'
Note that this is an innovation that the Celtic language share not only with the Italic languages, but also with Venetic. Also note that, of course, the P-Celtic languages essentially 'hidden' that feature because they have shifted *kw to *p, which is why the word is 'Pimpetos' in Gaulish, 'Pump' in Welsh and 'Pemp' in Breton.
Another common Italo-Celtic shape is the existence of -ī as a genitive form:
Latin - Dominus; Dominī
Gaulish - Mapos; Mapī
Archaic Irish - Maqqos; Maqqī
(it's interesting that this is attested in Archaic Irish, because this is generally a language, despite attested very late, that is rather close to Proto-Celtic)
There are also common words for gold, silver and tin in the Italic and Celtic languages, but not for iron. Conversely, the word for iron is shared between the Celtic and the Germanic languages, which matches the archaeological fact that iron working arrived in the (Proto-Germanic) Jastorf Culture through contact with the (Celtic) Hallstatt Culture.
It's really one of the main problems which I have with the Atlantic hypothesis is really that it somehow assumes the Celtic languages come out of thin air, woefully ignoring the interrelationship they have with the other branches of Indo-European. If the Celtic languages were really completely distinct from other branches of IE. Such can be seen with the Anatolian languages, which by the way, have retained some features of PIE lost elsewhere, such as the laryngal sounds.
I think you are wrong here. As I said, if you look at the distribution of R1a in western Europe, there are peaks in Auvergne and Cantabria. Are these coincidential? I don't really think they are. As for the connection between R1a and the IE languages I think the distribution in Eastern Europe is fairly compelling. For one, R1a has been found (exclusively) in graves of the Corded Ware Culture, and the distribution of R1a in Central-Eastern Europe (including Scandinavia) matches very well to the extend of Corded Ware. There is also a number of shared vocabulary shared by Germanic and Balto-Slavic languages which predates the Centum/Satem split (for example the word for gold). The Corded Ware Culture is very much fitting exactly these conditions for a common pattern.
The high percentages of Y-R1b-L21andwithin it L222 in W-Norway is intriguing me -
I find it very surprising slaves malesof Vikings ages should have been sent to Norway in so a huge numberand allowed to reproduce themselves (with the help of some females, Iconcede) to produce a 12-15% of some norwegian districts !
So I look for other events to explainit : Bronze Age ? The maritime distribution of Y-R-L21 inNW-Europe and its
apparent today association(statistical) with mt-H1 and H3, frequent from Iberia toScandinavia-Finland push me to believe in a possible old enoughsettling in W-Europe ; not to say R1b-L21 came from somewherewith mt-H1+H3, but that it found it at some stage and moved with somebearers of it between some places after that... - I'm not sure yetL21 were not in W-Europe before Bronze (we have almost no Y-DNA inNW-Europe between middle Neolithic and metal ages,
do correct meif I mistake ; – If I red well even L222 would not beborn in Ireland but only knew a founder effect in this isle (thissaid because the mtH1+H3 are not so dense in Ireland of today) -
I know some scholars said mtDNA had alot of chances to be partly natural pressure selected but at littlescale of time it does not seem to drastic, because H1+H3 seem strongin countries as far away the one from the other as Iberia andFinland... -
concerning Celts and Y-R1a I'm temptedto associate western R1a to Corded I consider as a proto-satemspeaking group – I see the Celts evolving in regions were BBs had astrong influence
where later were found rathermore Y-R1b and maybe Y-I2a2 (ex I1c ex I2b), without I think Celts =BBs at all ; I think BBs evolved with time, from a wellorganized federation of litlle groups of metals prospectors (I agreehere with the old traditional view) from central-south-east Europe(Croatia?) ultimately from South the Black Sea ? (according totheories concerning the bronze melting technics replacing the oldways of Balkans-Carpathians) – their skills permitted rapidprogresses in the regions they setlled – BBs I think had enoughskills and selfdefense power to be respected, not enough to invade orpopulate large countries – it's very possible that the last BBsaspects were more the result of accultured people, accultured byother accultured ones, were the first BBs had some descendants, moreand more crossed with the local populations as acculturation andassimilation were running on, these last in western Europe being forthe most celtic speaking ?...
to come back to Y-R1b-L21, they couldhave known two cultures, a non-I-Ean one and a I-Ean one, turned intoceltic early enough : the Qw- Celts ? I see them here a bitbefore S-Britain BBs settled from the Netherlands-Westphalen aboutthe 2500 BC -
& :'metal ages' is confusing : we can imagine metallurgists tookfoot in West before the all-around accepted datation for copper andbronze as collective well spred phenomenon in West) and BBs weremaybe not the first ones ?
- for Auvergne we have limited sample, with, I think, the weight of an industrial town as Clermont-Ferrand, so I have some mistrust in this lonesome survey – and do not forget East-Auvergne is in contact with Forez which was a terminal of Burgundian invaders as Lyonnais and Dauphiné (for Y-R1a) – I don't believe Celts had ever been strong for Y-R1a what does not exclude some rare SNP's... - but I confess the very low level of Y-I1 in Auvergne does not seem checking a northern germanic population : maybe are we building too much theories based on too little data here ?
Notice my pronostics are all bets andthat I want more ancient DNA to make a sound opinion, it's safer,evidently !!! – (my brain works better in analysis than insynthesis) – I need to tickle my brain to search possible ways ofarriving for R-L51 descendants in West – Mediterranea is maybe nota so ridiculous hypothesis for the P312 group ??? wellseparated from the ancestors of Y-R1b6U106 ? - some celticlegends involving E-Mediterranean and North Africa could have a partof truth ??? or they mix the true celtic history with the BBshistory, making a medley, what is not impossible at all knowing howlegends and myths work ?